RL+LLM Wiki

rl-llm-wiki's activity

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

lvwerra 

in rl-llm-wiki/knowledge-base about 19 hours ago

source: url:pretty-radio-b75.notion.site/DeepScaleR-Surpassing-O1-Preview-with-a-1-5B-Model-by-Scaling-RL-19681902c1468005bed8ca303013a4e2 — DeepScaleR 1.5B iterative-context-lengthening RL reproduction (open-repro)

3

#773 opened 4 days ago by

lvwerra

lvwerra 

in rl-llm-wiki/knowledge-base 1 day ago

source: url:interconnects.ai/p/summertime-outlook-o3s-novelty-coming — o3 novelty / RLVR search (speculation)

4

#728 opened 6 days ago by

lvwerra

lvwerra 

updated a bucket 1 day ago

rl-llm-wiki/rl-main-bucket

319 MB

lvwerra 

in rl-llm-wiki/knowledge-base 1 day ago

source: url:dwarkesh.com/p/john-schulman — Dwarkesh x John Schulman: pre-o1 reasoning/RL insider framing (transcript)

4

#775 opened 4 days ago by

lvwerra

source: url:nishtahir.com/notes-on-the-phi-4-reasoning-technical-paper — Phi-4-reasoning recipe reconstruction (Microsoft, secondary)

4

#765 opened 4 days ago by

lvwerra

source: url:newsletter.semianalysis.com/p/deepseek-debates — DeepSeek true training cost / $6M myth (speculation)

3

#740 opened 4 days ago by

lvwerra

lvwerra 

in rl-llm-wiki/knowledge-base 2 days ago

source: url:cnblogs.com/theseventhson/p/18725466 — Kimi k1.5 long-CoT RL reconstruction (CN, speculation)

4

#763 opened 4 days ago by

lvwerra

source: url:j-qi.medium.com/what-does-rl-improve-when-it-improves-llm-reasoning-2befa16c56e8 — RL = distribution shaping / distillation-vs-RL debate (speculation)

4

#772 opened 4 days ago by

lvwerra

lvwerra 

in rl-llm-wiki/knowledge-base 3 days ago

source: url:dwarkesh.com/p/dario-amodei-2 — Dwarkesh x Dario Amodei: RL scales log-linear like pretraining (transcript)

3

#778 opened 4 days ago by

lvwerra

source: url:dwarkesh.com/p/demis-hassabis — Dwarkesh x Demis Hassabis: AlphaZero-atop-LLMs, pre-o1 (transcript)

3

#779 opened 4 days ago by

lvwerra

source: url:cognitiverevolution.ai/the-rl-fine-tuning-playbook-coreweave-s-kyle-corbitt-on-grpo-rubrics-environments-reward-hacking — RL fine-tuning playbook: GRPO/rubrics/reward-hacking (practitioner transcript)

3

#784 opened 4 days ago by

lvwerra

source: url:interconnects.ai/p/rl-backlog-openais-many-rls-clarifying — OpenAI's many RLs / agentic (speculation)

4

#734 opened 6 days ago by

lvwerra

source: url:mechanize.work/blog/the-upcoming-gpt-3-moment-for-rl — RL environments as scaling substrate / GPT-3-moment thesis (forecast)

3

#785 opened 4 days ago by

lvwerra

lvwerra 

in rl-llm-wiki/knowledge-base 4 days ago

source: url:interconnects.ai/p/opening-the-black-box-of-character — Character training pipeline (speculation, paid/partial)

3

#732 opened 6 days ago by

lvwerra

source: url:pillumina.github.io/posts/aiinfra/02-slime — Zhipu GLM slime RL-infra source reconstruction (CN, speculation)

3

#764 opened 4 days ago by

lvwerra

source: url:interconnects.ai/p/thinking-searching-and-acting — Thinking/Searching/Acting (agentic speculation)

3

#731 opened 6 days ago by

lvwerra

source: url:yuanchaofa.com/post/kimi-k2-5-reading-notes — Kimi K2.5 PARL parallel-agent RL deep-read (CN, speculation)

3

#744 opened 4 days ago by

lvwerra

source: url:jasonwei.net/blog/asymmetry-of-verification-and-verifiers-law — Verifier's law / asymmetry of verification (speculation)

4

#739 opened 4 days ago by

lvwerra

source: url:cnblogs.com/volcengine-developer/articles/19070102 — veRL+ReTool tool-use RL reproduction / env+loss plumbing (CN, speculation)

3

#770 opened 4 days ago by

lvwerra

source: url:hkust-nlp.notion.site/simplerl-reason — SimpleRL-Zero 7B/8K R1-Zero reproduction / easy-to-hard (open-repro)

3

#774 opened 4 days ago by

lvwerra