Research map / Large language models / Alignment and preferences
OpenAssistant Conversations - Democratizing Large Language Model Alignment
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| SFT on OpenAssistant Conversations | Falcon-40B · LMEH | accuracy: 72.29 -> 74.04 | Helped |
| SFT on OpenAssistant Conversations | Pythia-12B · LMEH | accuracy: 60.33 -> 60.28 | No clear effect |
| RLHF instead of SFT only | LLaMA-30B OASST · Vicuna Elo | Elo: 979 -> 1068 | Helped |
| RLHF instead of SFT only | LLaMA-30B OASST · HumanEval | pass rate: 0.20 -> 0.15 | Hurt |
| mix OASST with other instruction data instead of top-1 threads only | Falcon-40B SFT · Vicuna Elo | Elo: 1192 -> 1053 | Hurt |
| mix OASST with other instruction data instead of top-1 threads only | Falcon-40B SFT · OpenAI Evals | score: 0.26 -> 0.44 | Helped |
Closest papers by meaning
- Exploring Intrinsic Alignments Within Text Corpus (AAAI 2025 · similarity 0.90)
- Aligner: Efficient Alignment by Learning to Correct (NeurIPS 2024 · similarity 0.90)
- Aligning Spoken Dialogue Models from User Interactions (ICML 2025 · similarity 0.89)
- Aligning to Thousands of Preferences via System Message Generalization (NeurIPS 2024 · similarity 0.89)
- The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning (ICLR 2024 · similarity 0.88)
- Your Weak LLM is Secretly a Strong Teacher for Alignment (ICLR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.