Search

Few direct matches — filled in with the latest updates.

Tag: #decomposition3 results

LEAD Breaks LLM No-Recovery Bottleneck

LEAD Breaks LLM No-Recovery Bottleneck

Long-horizon LLM execution is unstable due to a 'no-recovery bottleneck' from extreme decomposition and non-uniform errors. LEAD uses lookahead-enhanced atomic decomposition with short-horizon validation and overlapping rollouts to enable error correction. This allows o4-mini to solve Checkers Jumping puzzles up to n=13, surpassing n=11 limits.

ArXiv AIResearchMar 10#error-recovery#decomposition
Universe Segmentation Boosts Set Cover Optimization

Universe Segmentation Boosts Set Cover Optimization

Researchers introduce universe segmentability for decomposing Minimum Set Cover Problem (MSCP) instances into independent subproblems using union-find preprocessing. Subproblems are solved via GRASP metaheuristic with bit-level set representations for efficiency. Experiments demonstrate superior solution quality and scalability on benchmarks.

Evidence Alignment Bottleneck Exposed

Evidence Alignment Bottleneck Exposed

Decomposition boosts claim verification only with granular, sub-claim aligned evidence; repeated claim-level evidence degrades performance. Noisy sub-claim labels propagate errors unless using conservative abstention. New dataset features annotated evidence spans.

ArXiv AIResearchFeb 12#research#claim-verification#v1