Research map / Large language models / Language, safety and interpretability
Topological Active Inference for Task Disambiguation
Yangbo Wei, Zhen Huang, Shaoqiang Lu, Junhong Qian, Chen Wu, Lei He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| use persistent homology clustering instead of K-Means (ablation) | TAI · Ambi-Nav and Ambi-Code | success rate | Improved |
| increase sample size by 5x for Standard BED (ablation) | Standard BED · Ambi-Bench | success rate | No clear effect |
| use TAI topological structure with few samples (N=20) instead of Standard BED (ablation) | Standard BED · Ambi-Bench | success rate | Improved |
| replace Standard BED with TAI under variable renaming | Ambi-Code · Ambi-Code | success rate drop: -15.4% -> -0.5% | Improved |
| replace Standard BED with TAI under hybrid syntactic attack | Ambi-Code · Ambi-Code | success rate drop: -28.5% -> -1.2% | Improved |
| use TAI instead of Standard BED on a Qwen-0.6B model | Qwen-0.6B · Ambi-Bench | success rate (%): 15.2 -> 64.5 | Improved |
| use TAI instead of Standard BED on Qwen-14B | Qwen-14B · Ambi-Bench | success rate (%): 78.5 -> 99.2 | Improved |
| use TAI cluster-level soft belief update instead of candidate-level BED under combined noise | Ambi-Code · Ambi-Code, pflip=.20 punk=.20 | success rate (%): 38.6 -> 85.4 | Improved |
| use TAI instead of Standard BED on clean oracle | Ambi-Code · Ambi-Code | success rate (%): 75.0 -> 100 | Improved |
Closest papers by meaning
- Active Task Disambiguation with LLMs (ICLR 2025 · similarity 0.89)
- Persistent Topological Features in Large Language Models (ICML 2025 · similarity 0.87)
- Search for Truth from Reasoning: A Dynamic Representation Editing Framework for Steering LLM Trajectories (ICML 2026 · similarity 0.86)
- From Passive to Active Reasoning: Can Large Language Models Ask the Right Questions under Incomplete Information? (ICML 2025 · similarity 0.85)
- When Is Rank-1 Steering Cheap? Geometry, Granularity, and Budgeted Search (NeurIPS 2026 · similarity 0.85)
- Syntax vs. Semantics: How Transformers Learn Deep Dependencies (ICML 2026 · similarity 0.85)
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.
