GENIUS: Generative Fluid Intelligence Evaluation Suite
Paper • 2602.11144 • Published • 55
id stringlengths 3 3 | dimension stringclasses 1
value | task stringclasses 1
value | sub_dimension stringclasses 5
values | sub_sub_dimension null | context stringlengths 51 198 | instruction stringlengths 49 145 | rc_hint stringlengths 38 181 | vc_hint stringlengths 67 184 ⌀ | image_paths listlengths 1 4 |
|---|---|---|---|---|---|---|---|---|---|
002 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Smith likes the spatial relationship between the two objects in this image <image:space_0>. | Modify <image:space_1> to the spatial relationship that Smith likes. | The generated image should show coffee and books arranged horizontally. | <image:space_1> The appearance of coffee and book, and the background in this picture should be preserved. | [
"implicit_pattern/images/space_0.png",
"implicit_pattern/images/space_1.png"
] |
003 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | John dislikes the upper object in this image <image:space_9> but likes the lower object. John likes the upper object in this image <image:space_5> but dislikes the lower object. | Generate an image containing the objects that John would like. | The generated image should have the stream at the bottom and the clouds at the top. | null | [
"implicit_pattern/images/space_9.png",
"implicit_pattern/images/space_5.png"
] |
006 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | I like the spatial structure in this image <image:space_22>, but I don't like the spatial structure in this image <image:space_23>. | Please adjust this image <image:space_20> to make the objects within it conform to my preferred spatial structure. | The generated image should consist of five identical stones arranged at the five vertices of a pentagram. | <image:space_20> The size and style of the stone cannot be changed, and the entire background should be preserved. | [
"implicit_pattern/images/space_22.png",
"implicit_pattern/images/space_23.png",
"implicit_pattern/images/space_20.png"
] |
007 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Peter likes the right object's color in this image <image:space_0>, but dislikes the left object's color in this image <image:space_0> | Create a variant of <image:space_0> to make Peter like it. | The generated image should be the pineapple on the left turned yellow, and there was a sailboat on the right. | <image:space_0> The appearance of pineapple and sailboat should be preserved. The environment and background of this image should be preserved. | [
"implicit_pattern/images/space_0.png"
] |
008 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Britney likes the spatial structure in this image <image:space_23>, but she doesn't like the spatial structure in this image <image:space_22>. | Please adjust this image <image:space_20> to make the objects within it conform to Britney's preferred spatial structure. | The generated image should consist of five identical stones arranged diagonally. | <image:space_20> The size and style of the stone cannot be changed, and the entire background should be preserved. | [
"implicit_pattern/images/space_23.png",
"implicit_pattern/images/space_22.png",
"implicit_pattern/images/space_20.png"
] |
009 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Britney likes the spatial structure in this image <image:space_23>, but she doesn't like the spatial structure in this image <image:space_20>. | Please adjust this image <image:space_21> to make the objects within it conform to Britney's preferred spatial structure. | The generated image should consist of five identical stones arranged diagonally. | <image:space_21> The size and style of the stone cannot be changed, and the entire background should be preserved. | [
"implicit_pattern/images/space_23.png",
"implicit_pattern/images/space_20.png",
"implicit_pattern/images/space_21.png"
] |
010 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | I like the spatial structural relationship between the two objects in this image <image:space_0>. I dislike the spatial structural relationship between the two objects in this image <image:space_7>. | Create a variant of <image:space_7> and adjust it to the spatial structural relationship that I prefer. | The generated image should show a bird next to its nest. | <image:space_7> The bird's nest, the bird, and the background in this picture should be preserved. | [
"implicit_pattern/images/space_0.png",
"implicit_pattern/images/space_7.png"
] |
011 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | I like the inner object but dislike the outer object in this image <image:space_7>. I dislike the inner object but like the outer object in this image <image:space_2>. | Create an image that both objects are ones I like. | The generated image should show a small bird inside a wicker basket. | null | [
"implicit_pattern/images/space_7.png",
"implicit_pattern/images/space_2.png"
] |
012 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | John really likes the spatial structure in this image <image:space_24>, but he doesn't like the spatial structure in this image <image:space_25>. | Please adjust this image <image:space_26> to make the objects within it conform to John's preferred spatial structure. | The generated image should be: several small balls placed haphazardly around. | <image:space_26> The size and style of the balls cannot be changed. | [
"implicit_pattern/images/space_24.png",
"implicit_pattern/images/space_25.png",
"implicit_pattern/images/space_26.png"
] |
013 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | John really likes the spatial structure in this image <image:space_25>, but he doesn't like the spatial structure in this image <image:space_24>. | Please adjust this image <image:space_27> to make the objects within it conform to John's preferred spatial structure. | The generated image should be: several small balls arranged neatly in rows and columns. | <image:space_27> The size and style of the balls cannot be changed. | [
"implicit_pattern/images/space_25.png",
"implicit_pattern/images/space_24.png",
"implicit_pattern/images/space_27.png"
] |
015 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | John really likes the spatial structure in this image <image:space_28>. | Please adjust this image <image:space_29> to make the objects within it conform to John's preferred spatial structure. | The generated image should be: cookies arranged neatly from bottom to top and from largest to smallest. | <image:space_27> The size and style of the cookies cannot be changed. | [
"implicit_pattern/images/space_28.png",
"implicit_pattern/images/space_29.png",
"implicit_pattern/images/space_27.png"
] |
016 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Britney likes the spatial structure in this image <image:space_22>. | Please adjust this image <image:space_20> to make the objects within it conform to Britney's preferred spatial structure. | The generated image should consist of five identical stones arranged at the five vertices of a pentagram. | <image:space_20> The size and style of the stone cannot be changed. | [
"implicit_pattern/images/space_22.png",
"implicit_pattern/images/space_20.png"
] |
017 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Jenny thinks that the objects inside this image <image:space_7> are very interesting. | Create a variant of <image:space_7>, Jenny wants the interesting thing to be outside. | The generated image should show a bird outside its nest. | <image:space_7> The appearance of the bird and its nest should be preserved.
<image:space_7> The environment and background of this image should be preserved. | [
"implicit_pattern/images/space_7.png"
] |
018 | Implicit Pattern Induction | Implicit Pattern Generation | Spatial Relationship | null | Steve dislikes the inner object in this image <image:space_2>. | Create a variant of <image:space_2> to make Steve satisfied. | The generated image should contain only one empty wicker basket. | <image:space_2> The appearance of the box should be preserved.
<image:space_2> The environment and background of this image should be preserved. | [
"implicit_pattern/images/space_2.png"
] |
021 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Jack prefers the style in this image <image:overallstyle_0>. | Redraw <image:overallstyle_6> by adopting the specific style that Jack prefers. | The generated image should be a traditional Chinese meticulous painting. | <image:overallstyle_6> The appearance of the man and the appearance of background of this image should be preserved. | [
"implicit_pattern/images/overallstyle_0.png",
"implicit_pattern/images/overallstyle_6.png"
] |
022 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | I don't love the style in this image <image:overallstyle_1>, but I find this image <image:overallstyle_3> engaging. | Modify <image:overallstyle_4> to match the aesthetic preferred. | The generated image should be a simple, oil painting-style picture of a red car driving on the road. | <image:overallstyle_4> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_1.png",
"implicit_pattern/images/overallstyle_3.png",
"implicit_pattern/images/overallstyle_4.png"
] |
023 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Wang enjoys this image <image:overallstyle_2> but dislikes the style in this image <image:overallstyle_17>. | Apply the artistic style that Wang likes to <image:overallstyle_1>. | The generated image should be a steampunk-style street lined with tall buildings on both sides. | <image:overallstyle_1> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_2.png",
"implicit_pattern/images/overallstyle_17.png",
"implicit_pattern/images/overallstyle_1.png"
] |
025 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Sarah prefers this image <image:overallstyle_4> over the style in this image <image:overallstyle_12>. | Draw a picture of woman and a man using Sarah's favored style. | The generated image should be a picture of woman and a man in the extremely simple style. | null | [
"implicit_pattern/images/overallstyle_4.png",
"implicit_pattern/images/overallstyle_12.png"
] |
027 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Li loves the style in this image <image:overallstyle_6> and finds this image <image:overallstyle_7> unappealing. | Draw a picture of a man to mirror the style that Li love. | The generated image should be a color comic book version of a man. | null | [
"implicit_pattern/images/overallstyle_6.png",
"implicit_pattern/images/overallstyle_7.png"
] |
029 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_20> and finds this image <image:overallstyle_21> lack of aesthetic appeal. | Adapt <image:overallstyle_13> into the specific style that the artist loves. | The generated image should be: a warrior fighting a dragon, drawn entirely with crayons. | <image:overallstyle_13> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_20.png",
"implicit_pattern/images/overallstyle_21.png",
"implicit_pattern/images/overallstyle_13.png"
] |
031 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_22> and finds this image <image:overallstyle_21> lack of aesthetic appeal. | Adapt <image:overallstyle_25> into the specific style that the artist loves. | The generated image should consist of the sun, trees, mountains, and rivers; the overall style is Van Gogh's abstract art. | <image:overallstyle_25> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_22.png",
"implicit_pattern/images/overallstyle_21.png",
"implicit_pattern/images/overallstyle_25.png"
] |
032 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_24> and finds this image <image:overallstyle_26> lack of aesthetic appeal. | Adapt <image:overallstyle_23> into the specific style that the artist loves. | The generated image should be: a pop art style image: a man sitting by the window holding a cup. | <image:overallstyle_23> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_24.png",
"implicit_pattern/images/overallstyle_26.png",
"implicit_pattern/images/overallstyle_23.png"
] |
033 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | David loves this image <image:overallstyle_12> and dislikes the look of this image <image:overallstyle_7>. | Reconstruct <image:overallstyle_4> using the style that David enjoys. | The generated image should be a neon-lit style image of a red car driving on the road | <image:overallstyle_4> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_12.png",
"implicit_pattern/images/overallstyle_7.png",
"implicit_pattern/images/overallstyle_4.png"
] |
034 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_26> and finds this image <image:overallstyle_24> lack of aesthetic appeal. | Adapt <image:overallstyle_23> into the specific style that the artist loves. | The generated image should be: a collage art style image (combining different materials such as photos, clippings, and fabrics): an image of a man sitting by a window holding a cup. | <image:overallstyle_23> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_26.png",
"implicit_pattern/images/overallstyle_24.png",
"implicit_pattern/images/overallstyle_23.png"
] |
035 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_21> and finds this image <image:overallstyle_22> lack of aesthetic appeal. | Adapt <image:overallstyle_25> into the specific style that the artist loves. | The generated image should be: a mural-style image featuring a sun, trees, mountains, and a river. | <image:overallstyle_25> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_21.png",
"implicit_pattern/images/overallstyle_22.png",
"implicit_pattern/images/overallstyle_25.png"
] |
037 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Anna prefers the style in this image <image:overallstyle_2> and dislikes this image <image:overallstyle_4>. | Apply the visual style that matches Anna's taste to <image:overallstyle_12>. | The generated image should be a steampunk-style style image of a car driving on the road. | <image:overallstyle_12> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_2.png",
"implicit_pattern/images/overallstyle_4.png",
"implicit_pattern/images/overallstyle_12.png"
] |
038 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_22> and finds this image <image:overallstyle_21> lack of aesthetic appeal. | Adapt <image:overallstyle_24> into the specific style that the artist loves. | The generated image should be: a Van Gogh abstract art style image: a crying woman with a blue face and blonde hair. | <image:overallstyle_24> The woman's facial features, appearance, face and hair color, and the background must not be altered; the entire background should be preserved. | [
"implicit_pattern/images/overallstyle_22.png",
"implicit_pattern/images/overallstyle_21.png",
"implicit_pattern/images/overallstyle_24.png"
] |
039 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | Michael prefers this image <image:overallstyle_14> to the style in this image <image:overallstyle_0>. | Re-draw <image:overallstyle_9> in the spiritual style that matches Michael's taste. | The generated image should be in sketch style: a pine tree grows on a rock, and there is a white crane. | <image:overallstyle_9> All items displayed should remain unchanged, only the painting style should be altered. | [
"implicit_pattern/images/overallstyle_14.png",
"implicit_pattern/images/overallstyle_0.png",
"implicit_pattern/images/overallstyle_9.png"
] |
040 | Implicit Pattern Induction | Implicit Pattern Generation | Overall Style | null | The artist is fascinated by this image <image:overallstyle_20> and finds this image <image:overallstyle_21> lack of aesthetic appeal. | Adapt <image:overallstyle_24> into the specific style that the artist loves. | The generated image should be: a crayon-style image (the drawing must be done with crayons): a crying woman with a blue face and blonde hair. | <image:overallstyle_24> The woman's facial features, appearance, face and hair color, and the background must not be altered; the entire background should be preserved. | [
"implicit_pattern/images/overallstyle_20.png",
"implicit_pattern/images/overallstyle_21.png",
"implicit_pattern/images/overallstyle_24.png"
] |
041 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Jack likes the texture in this image <image:visualfeature_0>, but dislikes the texture of this image <image:visualfeature_15>. | Draw a picture of bag that align with Jack's preferences. | The generated image should be a green leather bag. | null | [
"implicit_pattern/images/visualfeature_0.png",
"implicit_pattern/images/visualfeature_15.png"
] |
042 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I prefer the hair in this image <image:visualfeature_3> and dislike the hair in this image <image:visualfeature_16>. | Draw a girl with beautiful hair to match the style I adore. | The generated image should be of a girl with thick, neat black hair. | null | [
"implicit_pattern/images/visualfeature_3.png",
"implicit_pattern/images/visualfeature_16.png"
] |
043 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Sarah appreciates the texture in this image <image:visualfeature_11>, but dislikes the texture in this image <image:visualfeature_19>. | Remake <image:visualfeature_2> based on Sarah's preferences. | The generated image should be a black and white checkered bowl. | <image:visualfeature_2> The environment and background of this image should be preserved. | [
"implicit_pattern/images/visualfeature_11.png",
"implicit_pattern/images/visualfeature_19.png",
"implicit_pattern/images/visualfeature_2.png"
] |
045 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Lee likes the pattern on the clothes in this image <image:visualfeature_20>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_21>. | Please adjust this image <image:visualfeature_23>, replacing what Mr. Lee dislikes with what he likes. | The generated image should be: a T-shirt with two cartoon pig patterns. | <image:visualfeature_23> The colors and styles of the clothing should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/visualfeature_20.png",
"implicit_pattern/images/visualfeature_21.png",
"implicit_pattern/images/visualfeature_23.png"
] |
046 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I am more fond of the surface texture in this image <image:visualfeature_6> and dislike the surface texture in this image <image:visualfeature_7>. | Draw a box to present the surface texture I like. | The generated image should be a box with a tree bark exterior like the bark of an old tree. | null | [
"implicit_pattern/images/visualfeature_6.png",
"implicit_pattern/images/visualfeature_7.png"
] |
047 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Elena prefers the patterns of this image <image:visualfeature_5> and dislikes <image:visualfeature_8>. | Update <image:visualfeature_15> to match Elena's specific preferences. | The generated image should be: a bag delicately embroidered with flowers and leaves. | <image:visualfeature_15> The shape and background of this image should be preserved. | [
"implicit_pattern/images/visualfeature_5.png",
"implicit_pattern/images/visualfeature_8.png",
"implicit_pattern/images/visualfeature_15.png"
] |
048 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I like the hair style in this image <image:visualfeature_16> and dislike the hair features in this image <image:visualfeature_10>. | Remake <image:visualfeature_3> to align with the preferences I displayed. | The generated image should be a girl with blonde curly hair. | <image:visualfeature_16> The hair style in this image should be preserved.
<image:visualfeature_15> The background of this image should be preserved. | [
"implicit_pattern/images/visualfeature_16.png",
"implicit_pattern/images/visualfeature_10.png",
"implicit_pattern/images/visualfeature_3.png",
"implicit_pattern/images/visualfeature_15.png"
] |
049 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | David appreciates the pattern in this image <image:visualfeature_13>, but dislikes the pattern in this image <image:visualfeature_15>. | Adjust <image:visualfeature_5> to the pattern David favors. | The generated image should be a sofa embroidered with a golden phoenix. | <image:visualfeature_13> The golden phoenix pattern in this image should be strictly preserved.
<image:visualfeature_5> The background of this image should be strictly preserved. | [
"implicit_pattern/images/visualfeature_13.png",
"implicit_pattern/images/visualfeature_15.png",
"implicit_pattern/images/visualfeature_5.png"
] |
050 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Lei likes the pattern on the clothes in this image <image:visualfeature_21>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_20>. | Please adjust this image <image:visualfeature_23>, replacing what Mr. Lei dislikes with what he likes. | The generated image should be: a T-shirt with two cartoon gray cat patterns. | <image:visualfeature_23> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_21.png",
"implicit_pattern/images/visualfeature_20.png",
"implicit_pattern/images/visualfeature_23.png"
] |
051 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Lei likes the pattern on the clothes in this image <image:visualfeature_20>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_21>. | Please adjust this image <image:visualfeature_22>, replacing what Mr. Lei dislikes with what he likes. | The generated image should be: a T-shirt with a cartoon pig pattern on the left and a cartoon dog pattern on the right. | <image:visualfeature_22> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_20.png",
"implicit_pattern/images/visualfeature_21.png",
"implicit_pattern/images/visualfeature_22.png"
] |
052 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Wang likes the pattern on the clothes in this image <image:visualfeature_25>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_26>. | Please adjust this image <image:visualfeature_28>, replacing what Mr. Wang dislikes with what he likes. | The generated image should be: a T-shirt with with two cartoon dog patterns. | <image:visualfeature_28> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_25.png",
"implicit_pattern/images/visualfeature_26.png",
"implicit_pattern/images/visualfeature_28.png"
] |
053 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Tom favors the hair style in <image:visualfeature_3> and is dissatisfied with <image:visualfeature_10>. | Redesign the <image:visualfeature_16> to match Tom's preference. | The generated image should be of a girl with thick, neat black hair. | <image:visualfeature_3> The hair style in this image should be preserved.
<image:visualfeature_16> The background of this image should be preserved. | [
"implicit_pattern/images/visualfeature_3.png",
"implicit_pattern/images/visualfeature_10.png",
"implicit_pattern/images/visualfeature_16.png"
] |
054 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Wang likes the pattern on the clothes in this image <image:visualfeature_26>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_25>. | Please adjust this image <image:visualfeature_28>, replacing what Mr. Wang dislikes with what he likes. | The generated image should be: a T-shirt with two cartoon cat patterns. | <image:visualfeature_28> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_26.png",
"implicit_pattern/images/visualfeature_25.png",
"implicit_pattern/images/visualfeature_28.png"
] |
055 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Wang likes the pattern on the clothes in this image <image:visualfeature_26>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_25>. | Please adjust this image <image:visualfeature_24>, replacing what Mr. Wang dislikes with what he likes. | The generated image should be: a T-shirt with a cartoon cat pattern on the left and a cartoon pig pattern on the right. | <image:visualfeature_24> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_26.png",
"implicit_pattern/images/visualfeature_25.png",
"implicit_pattern/images/visualfeature_24.png"
] |
056 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | Mr. Wang likes the pattern on the clothes in this image <image:visualfeature_26>, but he doesn't like the pattern on the clothes in this image <image:visualfeature_27>. | Please adjust this image <image:visualfeature_29>, replacing what Mr. Wang dislikes with what he likes. | The generated image should be: a T-shirt with two cartoon cat patterns. | <image:visualfeature_29> The colors and styles of the clothing should be strictly preserved, and the entire background should be strictly preserved. | [
"implicit_pattern/images/visualfeature_26.png",
"implicit_pattern/images/visualfeature_27.png",
"implicit_pattern/images/visualfeature_29.png"
] |
058 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I like the pattern on the vase in this image <image:visualfeature_30>, but I don't like the pattern on the scroll in this image <image:visualfeature_34>. | Convert <image:visualfeature_35> according to my aesthetic preference. | The generated image should be: a Chinese classical ink painting scroll depicting flowers. | <image:visualfeature_35> The scroll's style should be preserved, and the entire background should be retained. | [
"implicit_pattern/images/visualfeature_30.png",
"implicit_pattern/images/visualfeature_34.png",
"implicit_pattern/images/visualfeature_35.png"
] |
059 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I like the pattern on the vase in this image <image:visualfeature_31>, but I don't like the pattern on the scroll in this image <image:visualfeature_35>. | Convert <image:visualfeature_33> according to my aesthetic preference. | The generated image should be: a classical Chinese ink painting scroll depicting mountains and rivers. | <image:visualfeature_33> The scroll's style should be preserved, and the entire background should be retained. | [
"implicit_pattern/images/visualfeature_31.png",
"implicit_pattern/images/visualfeature_35.png",
"implicit_pattern/images/visualfeature_33.png"
] |
060 | Implicit Pattern Induction | Implicit Pattern Generation | Visual Feature | null | I like the pattern on the vase in this image <image:visualfeature_32>, but I don't like the pattern on the scroll in this image <image:visualfeature_33>. | Convert <image:visualfeature_34> according to my aesthetic preference. | The generated image should be: a classical Chinese ink painting scroll depicting fishes. | <image:visualfeature_34> The scroll's style should be preserved, and the entire background should be retained. | [
"implicit_pattern/images/visualfeature_32.png",
"implicit_pattern/images/visualfeature_33.png",
"implicit_pattern/images/visualfeature_34.png"
] |
061 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | I prefer the dog in this image <image:entity_0>. I dislike the dog in this image <image:entity_8>. | Modify <image:entity_16> by replacing the dog to match the aesthetic I prefer. | The generated image should be a golden retriever sitting on the ground in the street.. | <image:entity_0> The apperance of the golden retriever in this image should be preserved.
<image:entity_16> The background of this image should be preserved. | [
"implicit_pattern/images/entity_0.png",
"implicit_pattern/images/entity_8.png",
"implicit_pattern/images/entity_16.png"
] |
062 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Liam prefers the piano in this image <image:entity_7>. | Modify <image:entity_10> by replacing the main object with something to match the sophisticated style Liam prefers. | The generated image should be a piano. | <image:entity_7> The apperance of the piano in this image should be preserved.
<image:entity_10> The background of this image should be preserved. | [
"implicit_pattern/images/entity_7.png",
"implicit_pattern/images/entity_10.png"
] |
063 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Sophia prefers the cloth in this image <image:entity_10>. Sophia dislikes the cloth in this image <image:entity_2>. | Create an image with a deressed people to match the beauty Sophia prefers. | The generated image should be a person wearing a green coat. | null | [
"implicit_pattern/images/entity_10.png",
"implicit_pattern/images/entity_2.png"
] |
064 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | I prefer the flower in this image <image:entity_9>. I dislike the flower in this image <image:entity_3>. | Please draw me a picture: a large bouquet containing many of my favorite flowers. | The generated image should be a large bouquet containing many of phalaenopsis orchid. | <image:entity_9> The apperance of the flower in this image should be preserved. | [
"implicit_pattern/images/entity_9.png",
"implicit_pattern/images/entity_3.png"
] |
065 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Ethan prefers the female character in this image <image:entity_12>. Ethan dislikes the male character in this image <image:entity_18>. | Draw an image of male character to match the aesthetic Ethan prefers. | The generated image should be a cartoon schoolboy in a uniform. | null | [
"implicit_pattern/images/entity_12.png",
"implicit_pattern/images/entity_18.png"
] |
067 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | I really like the environment in this image <image:entity_20>. | Modify <image:entity_21> by replacing the background with the environment I like. | The generated image should show green grasslands on both sides of the road. | null | [
"implicit_pattern/images/entity_20.png",
"implicit_pattern/images/entity_21.png"
] |
068 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Noah prefers the electric guitar in this image <image:entity_15>. | Modify <image:entity_10> by replacing the main object with something to match the aesthetic Noah prefers. | The generated image should be a electric guitar. | <image:entity_15> The apperance of the electric guitar in this image should be preserved.
<image:entity_10> The background of this image should be preserved. | [
"implicit_pattern/images/entity_15.png",
"implicit_pattern/images/entity_10.png"
] |
069 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Emma prefers the style in this image <image:entity_4>. Emma dislikes the style in this image <image:entity_5>. | Create an image with a car running on the street to match the style Emma prefers. | The generated image should be a futuristic style car driving on the futuristic style street. | null | [
"implicit_pattern/images/entity_4.png",
"implicit_pattern/images/entity_5.png"
] |
070 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | I prefer the jacket in this image <image:entity_10>. | Modify <image:entity_7> by replacing the piano with something to match the my preference. | The generated image should be a green jacket. | <image:entity_10> The apperance of the green jacket in this image should be preserved.
<image:entity_7> The background of this image should be preserved. | [
"implicit_pattern/images/entity_10.png",
"implicit_pattern/images/entity_7.png"
] |
071 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Lucas likes the flower in this image <image:entity_9>. Lucas likes the bird in this image <image:entity_14>. | Draw a picture containing two favorite objects to match what Lucas likes. | The generated image should be a orchid with a flying bird. | <image:entity_9> The apperance of the flower in this image should be preserved.
<image:entity_14> The apperance of the bird in this image should be preserved. | [
"implicit_pattern/images/entity_9.png",
"implicit_pattern/images/entity_14.png"
] |
072 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Mia prefers the jacket in this image <image:entity_10>. | Modify <image:entity_15> by replacing the guitar with something to match the aesthetic Mia prefers. | The generated image should be a grenn jacket | <image:entity_15> The background of this image should be preserved. | [
"implicit_pattern/images/entity_10.png",
"implicit_pattern/images/entity_15.png"
] |
073 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Evan really likes the braided work in this image <image:entity_22>. The person on the right of this image <image:entity_23> is Evan. | Create a picture of Evan wearing his favorite clothes. | The generated image should be a little boy wearing a knitted hat. | <image:entity_23> The apperance of the boy in this image should be preserved.
<image:entity_22> The apperance of the knitted hat in this image should be preserved. | [
"implicit_pattern/images/entity_22.png",
"implicit_pattern/images/entity_23.png"
] |
074 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | James prefers the plant in this image <image:entity_3>. James dislikes the animal in this image <image:entity_6>. | Draw a picture that includes a plant James likes and an animal James dislikes. | The generated image should be a rose and a peacock. | null | [
"implicit_pattern/images/entity_3.png",
"implicit_pattern/images/entity_6.png"
] |
075 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Isabella prefers the electric guitar in this image <image:entity_15>. | Modify <image:entity_2> by replacing the coat with the object that Isabella prefers. | The generated image should be a guitar placed on a chair. | <image:entity_2> The background of this image should be preserved.
<image:entity_15> The apperance of the electric guitar in this image should be preserved. | [
"implicit_pattern/images/entity_15.png",
"implicit_pattern/images/entity_2.png"
] |
076 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | I prefer the animal in this image <image:entity_0>. | Modify <image:entity_7> by replacing the piano to match what I prefer. | The generated image should be a Golden Retriever on the stage. | <image:entity_7> The background of this image should be preserved.
<image:entity_0> The apperance of the golden retreiver in this image should be preserved. | [
"implicit_pattern/images/entity_0.png",
"implicit_pattern/images/entity_7.png"
] |
077 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Alexander prefers the character in this image <image:entity_4>. | Modify <image:entity_12> by transforming the character into one that matches the style Alexander prefers. | The generated image should be a futuristic girl with a high-tech feel standing in a hallway. | <image:entity_12> The background of this image should be preserved. | [
"implicit_pattern/images/entity_4.png",
"implicit_pattern/images/entity_12.png"
] |
078 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Charlotte prefers the flower in this image <image:entity_9>. Charlotte dislikes the flower in this image <image:entity_3>. | Please draw me a picture: the vase on the left has a flower Charlotte prefers , and the vase on the right has a flower Charlotte doesn't prefers. | The generated image should be: the vase on the left has a orchid, and the vase on the right has a rose. | <image:entity_9> The apperance of the flower in this image should be preserved in the left.
<image:entity_3> The apperance of the flower in this image should be preserved in the right. | [
"implicit_pattern/images/entity_9.png",
"implicit_pattern/images/entity_3.png"
] |
079 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Wang likes the items on the table in this image <image:entity_24>. | Create a variant of <image:entity_25> and transform the table layout object into what Wang prefers. | The generated image should show a rose placed in a vase on a table. | <image:entity_25> The environment and background of this image should be preserved. | [
"implicit_pattern/images/entity_24.png",
"implicit_pattern/images/entity_25.png"
] |
080 | Implicit Pattern Induction | Implicit Pattern Generation | Entity | null | Daniel prefers the coat in this image <image:entity_2>. | Modify <image:entity_19> by replacing the dress to match Daniel's preference. | The generated image should be a beige trench coat displayed on a hanger. | <image:entity_19> The background of this image should be preserved.
<image:entity_2> The apperance of the trench coat in this image should be preserved. | [
"implicit_pattern/images/entity_2.png",
"implicit_pattern/images/entity_19.png"
] |
081 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_0>. I dislike the palette in this image <image:colorpattern_1>. | Please fill this image <image:colorpattern_12> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of many overlapping glaciers, showcasing the extreme cold color pattern. | <image:colorpattern_12> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_0.png",
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_12.png"
] |
082 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_1>. I dislike the palette in this image <image:colorpattern_2>. | Please fill this image <image:colorpattern_12> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of many overlapping, lush green mountains, showcasing a green color pattern. | <image:colorpattern_12> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_2.png",
"implicit_pattern/images/colorpattern_12.png"
] |
083 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_2>. I dislike the palette in this image <image:colorpattern_3>. | Please fill this image <image:colorpattern_12> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of overlapping desert mountains displaying a yellowish-brown color pattern. | <image:colorpattern_12> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_2.png",
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_12.png"
] |
084 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_7>. I dislike the palette in this image <image:colorpattern_0>. | Please fill this image <image:colorpattern_12> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of a volcano with many overlapping layers of lava, displaying a dark red color pattern. | <image:colorpattern_12> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_7.png",
"implicit_pattern/images/colorpattern_0.png",
"implicit_pattern/images/colorpattern_12.png"
] |
085 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_8>. I dislike the palette in this image <image:colorpattern_1>. | Please fill this image <image:colorpattern_12> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of overlapping mountains in an autumn wilderness, displaying a brown color pattern. | <image:colorpattern_12> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_8.png",
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_12.png"
] |
086 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_0>. I dislike the palette in this image <image:colorpattern_1>. | Please fill this image <image:colorpattern_13> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image with many ice and snow crystal clusters, showcasing the color pattern of extreme cold. | <image:colorpattern_13> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_0.png",
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_13.png"
] |
087 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_3>. I dislike the palette in this image <image:colorpattern_4>. | Please fill this image <image:colorpattern_13> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image of many ice cream-colored crystal clusters, showcasing the soft color pattern of ice cream. | <image:colorpattern_13> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_4.png",
"implicit_pattern/images/colorpattern_13.png"
] |
088 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_4>. I dislike the palette in this image <image:colorpattern_6>. | Please fill this image <image:colorpattern_13> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image of many red, blue, or purple neon-colored crystal clusters, showing the color pattern of neon lights. | <image:colorpattern_13> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_4.png",
"implicit_pattern/images/colorpattern_6.png",
"implicit_pattern/images/colorpattern_13.png"
] |
089 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_9>. I dislike the palette in this image <image:colorpattern_3>. | Please fill this image <image:colorpattern_13> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image with many highly saturated bright reds, blues, and yellows, featuring strong contrasting colors, showcasing the color pattern of Pop Art. | <image:colorpattern_13> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_9.png",
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_13.png"
] |
090 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_6>. I dislike the palette in this image <image:colorpattern_3>. | Please fill this image <image:colorpattern_13> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image with many highly saturated bright reds, blues, and yellows, featuring strong contrasting colors, showcasing the color pattern of Pop Art. | <image:colorpattern_13> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_6.png",
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_13.png"
] |
091 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_9>. I dislike the palette in this image <image:colorpattern_3>. | Please fill this image <image:colorpattern_14> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a nautilus shell with many highly saturated bright reds, blues, and yellows, showcasing a strong contrasting color pattern characteristic of Pop Art. | <image:colorpattern_14> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_9.png",
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_14.png"
] |
092 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_3>. I dislike the palette in this image <image:colorpattern_2>. | Please fill this image <image:colorpattern_14> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of a nautilus shell with many soft, creamy ice cream colors, showcasing the color pattern of ice cream. | <image:colorpattern_14> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_2.png",
"implicit_pattern/images/colorpattern_14.png"
] |
093 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_5>. I dislike the palette in this image <image:colorpattern_3>. | Please fill this image <image:colorpattern_14> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image of a nautilus shell with a deep-sea blue color, showcasing the color pattern of the deep sea. | <image:colorpattern_14> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_5.png",
"implicit_pattern/images/colorpattern_3.png",
"implicit_pattern/images/colorpattern_14.png"
] |
094 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_10>. I dislike the palette in this image <image:colorpattern_7>. | Please fill this image <image:colorpattern_14> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image of a nautilus shell with multiple extremely bright, vibrant, high-purity color contrasts, showcasing the color pattern of dopamine. | <image:colorpattern_14> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_10.png",
"implicit_pattern/images/colorpattern_7.png",
"implicit_pattern/images/colorpattern_14.png"
] |
095 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_11>. I dislike the palette in this image <image:colorpattern_10>. | Please fill this image <image:colorpattern_14> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a black and white image of a cratered nautilus shell, showing the color pattern of the lunar surface. | <image:colorpattern_14> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_11.png",
"implicit_pattern/images/colorpattern_10.png",
"implicit_pattern/images/colorpattern_14.png"
] |
096 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_1>. I dislike the palette in this image <image:colorpattern_0>. | Please fill this image <image:colorpattern_15> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of green birds, showcasing a vibrant color pattern. | <image:colorpattern_15> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_0.png",
"implicit_pattern/images/colorpattern_15.png"
] |
097 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_2>. I dislike the palette in this image <image:colorpattern_1>. | Please fill this image <image:colorpattern_15> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of a bird with desert yellow hues, showcasing the color pattern of the desert. | <image:colorpattern_15> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_2.png",
"implicit_pattern/images/colorpattern_1.png",
"implicit_pattern/images/colorpattern_15.png"
] |
098 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_8>. I dislike the palette in this image <image:colorpattern_2>. | Please fill this image <image:colorpattern_15> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of a bird with colors of maple red, caramel and dark brown, showcasing the color pattern of an autumn wilderness. | <image:colorpattern_15> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_8.png",
"implicit_pattern/images/colorpattern_2.png",
"implicit_pattern/images/colorpattern_15.png"
] |
099 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_9>. I dislike the palette in this image <image:colorpattern_10>. | Please fill this image <image:colorpattern_15> with my preferred palette, and show the characteristics of the palette. | The generated image should be: a picture of a bird with highly saturated bright red, blue, and yellow colors, showcasing the color pattern of Pop Art. | <image:colorpattern_15> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_9.png",
"implicit_pattern/images/colorpattern_10.png",
"implicit_pattern/images/colorpattern_15.png"
] |
100 | Implicit Pattern Induction | Implicit Pattern Generation | Palette | null | I prefer the palette in this image <image:colorpattern_10>. I dislike the palette in this image <image:colorpattern_9>. | Please fill this image <image:colorpattern_15> with my preferred palette, and show the characteristics of the palette. | The generated image should be: an image of a bird with a variety of extremely bright, vibrant, high-purity colors, showcasing the color pattern of dopamine. | <image:colorpattern_15> The lines and basic shape should be preserved, and the entire background should be preserved. | [
"implicit_pattern/images/colorpattern_10.png",
"implicit_pattern/images/colorpattern_9.png",
"implicit_pattern/images/colorpattern_15.png"
] |
Leaderboard
GENIUS evaluates every generated image on three complementary axes, each scored 0 (fail), 1 (partial), or 2 (perfect):
- Rule Compliance (RC): follows the newly defined rule, grounded by expert-written evaluation hints.
- Visual Consistency (VC): preserves required identities, objects, and contextual visual attributes.
- Aesthetic Quality (AQ): remains visually coherent and avoids obvious generation artifacts.
The Overall score uses the paper's weighted metric ratio RC:VC:AQ = 6:3.5:0.5. The primary Gemini score averages three independent judge runs per sample. Qwen2.5-VL-72B is an independent robustness judge: it assigns lower absolute scores overall, while preserving the main relative performance trends.
| Rank by Gemini | Model / Method | Type | Interleaved | Gemini-3-Pro Overall | Qwen2.5-VL-72B Overall |
|---|---|---|---|---|---|
| 1 | Nano Banana Pro | Proprietary | ✓ | 57.19 | 48.35 |
| 2 | SeeDream 4.5 | Proprietary | ✗ | 52.84 | 44.17 |
| 3 | Nano Banana | Proprietary | ✓ | 50.66 | 42.88 |
| 4 | GPT-Image | Proprietary | ✗ | 47.15 | 40.94 |
| 5 | Emu3.5-Image | Open source | ✗ | 36.67 | 28.80 |
| 6 | FLUX.2-dev | Open source | ✗ | 34.39 | 27.37 |
| 7 | Ours (Bagel + attention intervention) | Training-free method | ✓ | 32.92 | 23.91 |
| 8 | Qwen-Image | Open source | ✗ | 30.58 | 25.67 |
| 9 | Omini-Gen2 | Open source | ✗ | 27.87 | 21.12 |
| 10 | Bagel | Open source | ✓ | 26.74 | 18.97 |
| 11 | GLM-Image | Open source | ✗ | 24.71 | 17.45 |
| 12 | SeeDream 4.0 | Proprietary | ✗ | 21.26 | 17.74 |
| 13 | NextStep-1 | Open source | ✗ | 10.44 | 9.90 |
GENIUS leaderboard judged by Gemini-3-Pro:
What is GENIUS?
Most image-generation benchmarks primarily test crystallized intelligence: whether a model can retrieve and reproduce concepts learned during pre-training. GENIUS instead evaluates **Generative Fluid Intelligence (GFI)**—the ability to solve visual-generation problems whose rules are defined entirely by the immediate multimodal context.
GENIUS contains 510 expert-curated test cases with interleaved text and multi-image inputs. Removing either modality makes an instance unsolvable. The benchmark spans three core capabilities:
| GFI capability | What it tests | Tasks | Samples |
|---|---|---|---|
| 🧩 Implicit Pattern Induction | Infer unstated visual preferences and apply them to a new image | Implicit Pattern Generation | 86 |
| 📐 Ad-hoc Constraint Execution | Execute newly defined visual or symbolic rules | Symbolic Constraint Generation; Visual Constraint Generation | 213 |
| 🌍 Contextual Knowledge Adaptation | Follow contextual knowledge even when it conflicts with familiar semantics or physical priors | Prior-Conflicting Generation; Multi-Semantic Generation | 211 |
| Total | 3 dimensions · 5 tasks · 20 sub-tasks | 510 |
Dataset composition
| Viewer config | Dimension | Sub-tasks | Samples |
|---|---|---|---|
implicit_pattern |
Implicit Pattern Induction | Overall Style (15), Spatial Relationship (14), Visual Feature (18), Palette (20), Entity (19) | 86 |
symbolic_constraint |
Ad-hoc Constraint Execution | Operation Implementation (27), Visual Metaphor (29), Layout (30), Visual Feature (38), Instance Binding (29) | 153 |
visual_constraint |
Ad-hoc Constraint Execution | Simple Constraint (26), Complex Constraint (34) | 60 |
prior_conflicting |
Contextual Knowledge Adaptation | Abnormal Biological Growth (21), Gravity Anomaly (20), Abnormal Animal Behavior (20), Time Reversal (20), Weather Anomaly (20) | 101 |
multi_semantic |
Contextual Knowledge Adaptation | Noun Phrase (50), Verb Phrase (30), Adjective Phrase (30) | 110 |
Each record contains:
| Field | Description |
|---|---|
id |
Task-local identifier used to name the generated output |
dimension / task |
GFI capability dimension and benchmark task |
sub_dimension |
Fine-grained capability category |
sub_sub_dimension |
Additional fine-grained category used by some symbolic constraints |
context |
Interleaved context that establishes the new rule or pattern |
instruction |
Image-generation or image-editing request |
rc_hint |
Human-verified gold hint for Rule Compliance evaluation |
vc_hint |
Human-verified gold hint for Visual Consistency, when applicable |
image_paths |
Ordered, de-duplicated repository paths for images referenced by the record |
Image placeholders use the form <image:NAME> and resolve to CONFIG/images/NAME.png. For example, <image:space_0> in the implicit_pattern config refers to implicit_pattern/images/space_0.png.
Viewer representation. The five
viewer/*.jsonlfiles are generated from the canonical*/test_data.jsonannotations. They preserve all annotation text while normalizing the occasionally mixed string/list representation ofvc_hintinto a nullable string, giving the Dataset Viewer a stable schema. Rebuild them withpython scripts/build_viewer_data.pyafter changing the source annotations.
Quick start
Load any of the five task configs directly with 🤗 Datasets:
from datasets import load_dataset
dataset = load_dataset(
"HankYang428/GENIUS",
"implicit_pattern", # or symbolic_constraint, visual_constraint,
# prior_conflicting, multi_semantic
split="test",
)
print(dataset[0])
Download the full repository, including all reference images:
git lfs install
git clone https://huggingface.co/datasets/HankYang428/GENIUS
Alternative mirrors: Google Drive · Baidu Netdisk (password: iek1)
For model-output layout and evaluation commands, see the official code repository.
Intended use and limitations
GENIUS is designed for evaluating multimodal image-generation and image-editing systems on novel, context-defined problems. It is a compact evaluation suite, not a training corpus. Scores measure performance on the benchmark's curated rules and judge protocol; they should not be interpreted as a complete measure of general intelligence or image-generation quality. LMM-as-a-judge evaluation can retain model-specific biases, which is why the paper reports results from two independent judge families and uses expert-written gold hints.
License
GENIUS is released under the Creative Commons Attribution-NonCommercial 4.0 International License and is intended for non-commercial research use.
Citation
If GENIUS is useful in your research, please cite:
@misc{an2026geniusgenerativefluidintelligence,
title = {GENIUS: Generative Fluid Intelligence Evaluation Suite},
author = {Ruichuan An and Sihan Yang and Ziyu Guo and Wei Dai and Zijun Shen and Haodong Li and Renrui Zhang and Xinyu Wei and Guopeng Li and Wenshan Wu and Wentao Zhang},
year = {2026},
eprint = {2602.11144},
archivePrefix = {arXiv},
primaryClass = {cs.LG},
url = {https://arxiv.org/abs/2602.11144}
}
Contact
Questions and feedback are welcome via GitHub Issues or arctanxarc@gmail.com.
- Downloads last month
- 699