LIMA

anchor

Findings

• IC-1576Base and aligned LLMs share 77.7% of top-1 token predictions, with distribution shifts concentrated in stylistic tokens rather than knowledge content [context]
• IC-1577Base LLMs prompted with URiAL (3 restyled in-context examples + system prompt) match or surpass their SFT/RLHF-aligned counterparts on multi-aspect evaluation [context]
• IC-1577Base LLMs prompted with URiAL (3 restyled in-context examples + system prompt) match or surpass their SFT/RLHF-aligned counterparts on multi-aspect evaluation [source]
• IC-986Most LLMs lack tool usage awareness, with only ChatGPT exceeding 70% F1 in zero-shot evaluation [eval]