i

iliashum's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

commented a paper 9 days ago

Soft Instruction De-escalation Defense

Paper • 2510.21057 • Published 12 days ago • 3 •

1

commented a paper 14 days ago

Extracting alignment data in open models

Paper • 2510.18554 • Published 15 days ago • 7 •

3

commented a paper 21 days ago

SynthID-Image: Image watermarking at internet scale

Paper • 2510.09263 • Published 26 days ago • 1 •

2

commented a paper 22 days ago

The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections

Paper • 2510.09023 • Published 26 days ago • 8 •

2

commented a paper about 2 months ago

Reasoning Introduces New Poisoning Attacks Yet Makes Them More Complicated

Paper • 2509.05739 • Published Sep 6 • 2 •

3

commented a paper 4 months ago

Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities

Paper • 2507.06261 • Published Jul 7 • 63 •

4

commented 3 papers 5 months ago

Cascading Adversarial Bias from Injection to Distillation in Language Models

Paper • 2505.24842 • Published May 30 • 6 •

2

Architectural Backdoors for Within-Batch Data Stealing and Model Inference Manipulation

Paper • 2505.18323 • Published May 23 • 3 •

2

Strong Membership Inference Attacks on Massive Datasets and (Moderately) Large Language Models

Paper • 2505.18773 • Published May 24 • 7 •

2

commented 3 papers 6 months ago

Fixing 7,400 Bugs for 1$: Cheap Crash-Site Program Repair

Paper • 2505.13103 • Published May 19 • 6 •

2

Lessons from Defending Gemini Against Indirect Prompt Injections

Paper • 2505.14534 • Published May 20 • 8 •

2

Humans expect rationality and cooperation from LLM opponents in strategic games

Paper • 2505.11011 • Published May 16 • 5 •

2

commented a paper 7 months ago

Defeating Prompt Injections by Design

Paper • 2503.18813 • Published Mar 24 • 22 •

1

commented a paper 10 months ago

Trusted Machine Learning Models Unlock Private Inference for Problems Currently Infeasible with Cryptography

Paper • 2501.08970 • Published Jan 15 • 6 •

2

commented a paper 12 months ago

Hardware and Software Platform Inference

Paper • 2411.05197 • Published Nov 7, 2024 • 4 •

2

commented 3 papers about 1 year ago

Stealing User Prompts from Mixture of Experts

Paper • 2410.22884 • Published Oct 30, 2024 • 15 •

2

Measuring memorization through probabilistic discoverable extraction

Paper • 2410.19482 • Published Oct 25, 2024 • 4 •

2

Operationalizing Contextual Integrity in Privacy-Conscious Assistants

Paper • 2408.02373 • Published Aug 5, 2024 • 5 •

2

commented 2 papers over 1 year ago

A False Sense of Safety: Unsafe Information Leakage in 'Safe' AI Responses

Paper • 2407.02551 • Published Jul 2, 2024 • 9 •

1

UnUnlearning: Unlearning is not sufficient for content regulation in advanced generative AI

Paper • 2407.00106 • Published Jun 27, 2024 • 6 •

1