analiza aiqbench
This research introduces AdaMAST, a method to automatically generate compact, evidence-based taxonomies from agent execution traces without manual annotation. By organizing failures along system-level, role-specific, and domain-specific axes, the approach significantly boosts performance in agent search, runtime monitoring, and trajectory selection benchmarks. The work advances AI engineering by providing reusable diagnostic vocabularies that enhance feedback loops for autonomous agents beyond simple free-text reflection.
Fantastic Adaptive Taxonomies and How to Use Them
arXiv:2607.16387v1 Announce Type: cross Abstract: An agent system's execution traces record how it fails, and procedures that improve such a system without changing model weights (trajectory selection, prompt and workflow optimization, runtime monitoring) read these traces for feedback. Yet raw traces are a poor medium for accumulating that feedback: long, instance-specific, and lacking a stable vocabulary for recurring failures. We argue that an agent system should instead maintain an explicit representation of how it fails, induced from its own behavior and reusable wherever failure feedback is needed. AdaMAST builds this representation by converting a target system's traces into a compact, evidence-grounded failure taxonomy: named failure codes organized along three fixed axes (system-level, role-specific, and domain-specific), with every name, definition, and evidence pattern induced from the traces; no code is hand-authored, no trace human-annotated. The taxonomy is not merely a post-hoc diagnostic but a shared feedback interface, improving agents in three ways. In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form reflection on all five benchmarks we test. At runtime, taxonomy feedback raises SWE-agent's resolution on SWE-bench Verified Mini from 60% with free-text reflection to 70%, and improves Claude Code from 64.0% to 70.7% as a runtime skill. In trajectory selection, AdaMAST-Judge, a verifier built on the induced codes, improves best-of-5 accuracy on Terminal-Bench 2.0 by 8-15 points over Pass@1. The vocabulary itself is compact (an order-of-magnitude compression that preserves trace distinctions), human-faithful (matching expert failure annotations more closely than a hand-crafted reference vocabulary), and adaptive (taxonomies induced for different domains share few codes). Adaptive failure taxonomies close the loop between the traces agents produce and the procedures that improve them.
Authors:Mert Cemri, Andrei Cojocaru, Melissa Pan, Shu Liu, Shubham Agarwal, Alexander Krentsel, Jay Tang, Kannan Ramchandran, Joseph E. Gonzalez, Matei Zaharia, Alex Dimakis, Ion Stoica
Abstract:An agent system's execution traces record how it fails, and procedures that improve such a system without changing model weights (trajectory selection, prompt and workflow optimization, runtime monitoring) read these traces for feedback. Yet raw traces are a poor medium for accumulating that feedback: long, instance-specific, and lacking a stable vocabulary for recurring failures. We argue that an agent system should instead maintain an explicit representation of how it fails, induced from its own behavior and reusable wherever failure feedback is needed. AdaMAST builds this representation by converting a target system's traces into a compact, evidence-grounded failure taxonomy: named failure codes organized along three fixed axes (system-level, role-specific, and domain-specific), with every name, definition, and evidence pattern induced from the traces; no code is hand-authored, no trace human-annotated. The taxonomy is not merely a post-hoc diagnostic but a shared feedback interface, improving agents in three ways. In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form reflection on all five benchmarks we test. At runtime, taxonomy feedback raises SWE-agent's resolution on SWE-bench Verified Mini from 60% with free-text reflection to 70%, and improves Claude Code from 64.0% to 70.7% as a runtime skill. In trajectory selection, AdaMAST-Judge, a verifier built on the induced codes, improves best-of-5 accuracy on Terminal-Bench 2.0 by 8-15 points over Pass@1. The vocabulary itself is compact (an order-of-magnitude compression that preserves trace distinctions), human-faithful (matching expert failure annotations more closely than a hand-crafted reference vocabulary), and adaptive (taxonomies induced for different domains share few codes). Adaptive failure taxonomies close the loop between the traces agents produce and the procedures that improve them.
Software Engineering (cs.SE); Artificial Intelligence (cs.AI)
arXiv:2607.16387 [cs.SE]
(or arXiv:2607.16387v1 [cs.SE] for this version)
https://doi.org/10.48550/arXiv.2607.16387 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 312 cuvinte
Subiecte conexe
Presentation: Building Evals for AI Adoption: from Principles to Practice
InfoQ · 29 mai 2026
TACT: Taxonomy-Aligned Post-Training for Pedagogically Adaptive English Tutoring
arXiv · 5 aug. 2026
A Taxonomy of Cognitive Capability Gaps in Generative and Agentic AI
arXiv · 5 aug. 2026
A Six-Dimensional Taxonomy of Post-Training Adaptation Techniques with Applications in AI Governance
arXiv · 7 aug. 2026