Get Started
Research questionHow can in-context learning select demonstrations by flexible reasoning logic rather than surface similarity?Surface-similar examples can use different solution strategies, while exact logical matching can miss valid variations in how problems are solved.
AI
Information Retrieval
Natural Language Processing
Reasoning
Latest papersRecent research connected to this question, newest first.SALA: Semantic-Aware Logical Alignment for Complex Reasoning in In-Context LearningThe source studies SALA, which induces task-specific reasoning operations and aligns their sequences in a continuous semantic space. Evidence covers four reasoning benchmarks and three LLMs, with analyses of operation induction and logical alignment.research paper · Sep 2, 2026
Related questions
How can language models perform complex logical reasoning without accumulating token-level errors?How should on-policy distillation select examples for data-efficient reasoning post-training?How can reinforcement learning post-training prioritize useful reasoning prompts as learning signals shift?How can reasoning models keep improving on open-ended agentic tasks as human supervision and reliable rewards recede?
Home
Topics
Search
Library