Leandro von Werra PRO

lvwerra's activity

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

New activity in rl-llm-wiki/knowledge-base 29 minutes ago

topic: reward-model-ensembles — deepen to the flagship bar (12.2KB → 16.2KB)

2

#321 opened about 1 hour ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base 31 minutes ago

topic: process-vs-outcome-rewards — add mechanism, design-space table, runnable trace-error check

#322 opened 31 minutes ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base about 1 hour ago

topic: human-preference-collection — deepen to the flagship bar (11.2KB → 16.7KB)

2

#320 opened about 2 hours ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base about 2 hours ago

topic: ai-feedback-data — deepen to the flagship bar (10.7KB → 18.5KB)

2

#318 opened about 2 hours ago by

lvwerra

topic: reasoning-emergence §5 — add the mechanism (cognitive behaviors + entropy collapse) to the created-vs-surfaced debate

2

#319 opened about 2 hours ago by

lvwerra

fix: eval-cluster consistency pass — add missing back-links (bidirectional navigation)

#317 opened about 2 hours ago by

lvwerra

topic: data-quality-and-filtering — deepen to the flagship bar (9.9KB → 17.3KB)

2

#316 opened about 2 hours ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base about 3 hours ago

topic: test-time-and-rl-interplay — deepen to the flagship bar (9.2KB → 16.8KB)

2

#315 opened about 3 hours ago by

lvwerra

topic: verifiable-rewards — deepen to the flagship bar (9.9KB → 20.9KB)

2

#314 opened about 3 hours ago by

lvwerra

fix: adversarial-robustness → deceptive-alignment reciprocal cross-link

2

#313 opened about 3 hours ago by

lvwerra

topic: evaluation/llm-as-judge — deep synthesis node (one mechanism, two masters: eval metric + training reward)

2

#311 opened about 4 hours ago by

lvwerra

fix: safety-cluster cross-links (open-problems↔deceptive-alignment↔adversarial-robustness)

2

#312 opened about 4 hours ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base about 4 hours ago

topic: safety-and-alignment/adversarial-robustness-and-jailbreaks

4

#309 opened about 5 hours ago by

lvwerra

topic: credit-granularity — fold VinePPO (the advantage-estimation facet of credit assignment)

2

#310 opened about 5 hours ago by

lvwerra

fix: deepen scalable-oversight §4 with empirical debate, easy→hard, prover-verifier (absorbs 3 orphan sources)

4

#288 opened 3 days ago by

lvwerra

New activity in rl-llm-wiki/knowledge-base about 5 hours ago

topic: safety-and-alignment/deceptive-alignment — deep node (inner misalignment & how RL interacts)

2

#308 opened about 5 hours ago by

lvwerra

topic: algorithms/credit-granularity-in-preference-optimization — deep synthesis of the credit-granularity axis

3

#307 opened about 5 hours ago by

lvwerra

fix: capability-and-safety-benchmarks — link to agentic-benchmarks deep child (hub bidirectional link)

2

#306 opened about 6 hours ago by

lvwerra

fix: capability-and-safety-benchmarks — link to agentic-benchmarks deep child (hub bidirectional link)

2

#306 opened about 6 hours ago by

lvwerra

topic: safety-and-alignment/adversarial-robustness-and-jailbreaks

4

#309 opened about 5 hours ago by

lvwerra