REFLECT uses intervention-supported diagnosis and outcome verification to achieve the highest error localization accuracy among same-auditor methods on four multi-hop reasoning benchmarks.
arXiv preprint arXiv:2512.06749 , year=
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 7roles
background 1polarities
background 1representative citing papers
Refploit repairs code-agent trajectories for Java exploit reproduction via differential validation and focused recovery constraints, achieving 80.2% success on 172 references with 64.3% relative improvement.
POIROT protocol repurposes agents in LLM multi-agent systems as an internal diagnostic layer for failure detection, outperforming single-LLM evaluators with gains that increase with complexity, agent count, and fault types.
Introduces EPC-AW to mitigate epistemic miscalibration in LLM multi-agent planning via consistency-based selection and refinement, reporting 9.75% average success improvement.
Token-level contrastive attribution yields informative signals for some LLM benchmark failures but is not universally applicable across datasets and models.
Insights Generator is a multi-agent system that produces evidence-backed insights from corpora of LLM agent traces and yields 30.4pp performance gains when humans apply the reports.
RGAO combines retrieval-based complexity assessment with a formal budget algebra to enable dynamic topology selection in multi-agent code generation with provable conservation.
citing papers explorer
-
REFLECT: Intervention-Supported Error Attribution for Silent Failures in LLM Agent Traces
REFLECT uses intervention-supported diagnosis and outcome verification to achieve the highest error localization accuracy among same-auditor methods on four multi-hop reasoning benchmarks.
-
Refploit: Facilitating Exploit Construction via Code-Agent Trajectory Repair
Refploit repairs code-agent trajectories for Java exploit reproduction via differential validation and focused recovery constraints, achieving 80.2% success on 172 references with 64.3% relative improvement.
-
POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems
POIROT protocol repurposes agents in LLM multi-agent systems as an internal diagnostic layer for failure detection, outperforming single-LLM evaluators with gains that increase with complexity, agent count, and fault types.
-
When Planning Fails Despite Correct Execution: On Epistemic Calibration for LLM-Based Multi-Agent Systems
Introduces EPC-AW to mitigate epistemic miscalibration in LLM multi-agent planning via consistency-based selection and refinement, reporting 9.75% average success improvement.
-
Contrastive Attribution in the Wild: An Interpretability Analysis of LLM Failures on Realistic Benchmarks
Token-level contrastive attribution yields informative signals for some LLM benchmark failures but is not universally applicable across datasets and models.
-
Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents
Insights Generator is a multi-agent system that produces evidence-backed insights from corpora of LLM agent traces and yields 30.4pp performance gains when humans apply the reports.
-
Retrieval-Conditioned Topology Selection with Provable Budget Conservation for Multi-Agent Code Generation
RGAO combines retrieval-based complexity assessment with a formal budget algebra to enable dynamic topology selection in multi-agent code generation with provable conservation.