arXiv AI

Skill-based Agentic Evaluation for Real-time Data Science Tasks

arXiv AI
Sep 28

CARGO: Context-Aware Retrieval-Gated Evaluation of Agentic AI in Production

CARGO is a framework for evaluating agentic AI systems in production that addresses the problem of reference-instance divergence (RID), where reference-based judges penalize correct answers that involve different entity identifiers. It treats retrieved references as procedural exemplars, grounds judgments in the live instance’s context, assigns a three-way status to claims, and gates evaluation by retrieval confidence. Using the CARGO-Bench diagnostic suite, CARGO eliminates false penalties and improves discrimination while revealing a limitation in detecting procedural corruptions.

By Mukul Chhabra, Shail Patel, Luigi Medrano
arXiv AI
Sep 7

TruthInsightBench: An Evidence-Grounded Benchmark for Automated Evaluation of Open-Ended Scientific Discovery Agents

TruthInsightBench is a new benchmark designed to evaluate automated scientific discovery agents by presenting them with 40 blind tasks drawn from peer‑reviewed studies across ten domains. Each task provides only a neutral objective and frozen data, withholding source conclusions, expected values, and analysis paths, forcing agents to determine which claim the data support. A fixed LLM‑based judge scores agents on evidentiary maturity across six dimensions, using 29 artifact‑grounded items, enabling fully automated, repeatable evaluation without human grading.

By Zhibo Yang, Chen Zhang, Yuewei Zhang, Hao Wang
arXiv AI
Jun 2

BADGER: Bridging Agentic and Deterministic Evaluation for Generative Enterprise Reasoning

arXiv:2606. 02109v1 Announce Type: new Abstract: Enterprise AI systems that translate natural language into SQL queries and orchestrate multi-step agentic reasoning pipelines require evaluation approaches fundamentally different from academic benchmarks.

By Shannon Serrao, Soumitra Chatterjee, Dorina Strori, Abhishek Sharma, Nathan Miller
arXiv AI
Jun 17

A Framework for Evaluating Agentic Skills at Scale

arXiv:2606. 17819v1 Announce Type: cross Abstract: Agent skills -- structured, reusable knowledge artifacts that augment LLM agent capabilities -- have been rapidly adopted in industry, yet their cross-domain impact and use across commercial and open-source models remain under-studied, and no reusable methodology exists for evaluating an individual skill.

By Maksim Shaposhnikov, Nicolas Fortuin, Simon Stipcich, Maria I. Gorinova, Amy Heineike, Rob Willoughby
arXiv AI
Jun 11

Can AI Agents Synthesize Scientific Conclusions?

arXiv:2606. 11337v1 Announce Type: new Abstract: Scientific AI agents increasingly retrieve evidence, reason across sources, and synthesize conclusions used in consequential decisions.

By Hayoung Jung, Pedro Viana Diniz, Jos\'e Reinaldo Corr\^ea Roveda, Abner Fernandes da Silva, Haeun Jung, Enoch Tsai, Aleksandra Korolova, Manoel Horta Ribeiro
arXiv AI
Oct 12

SWE-Critix: Post-Training Generalist Verifiers for Assessing Coding-Agent Trajectories

SWE-Critix is a post‑training pipeline that creates generalist verifiers for coding‑agent trajectories. It uses supervised fine‑tuning followed by reinforcement learning to train 30B models on CodeForge‑Preview data, enabling the verifier to predict completion labels and generate natural‑language explanations. The approach raises classification accuracy from 57% to 78% on a held‑out test set.

By Jiajun Hu, Chun Yong Chong
arXiv AI
Oct 21

Navalia: Advancing Cowork Agents through Verifiable Task Synthesis and Long-Horizon Post-Training

Navalia is a framework that advances cowork agents by combining verifiable task synthesis with long‑horizon post‑training. It selects supporting evidence from corporate financial reports, builds executable computation chains, and uses independent LLM review and programmatic checks to ensure evidence consistency and answer correctness. The verified tasks are then used for trajectory collection, supervised fine‑tuning, and reinforcement learning, yielding consistent performance gains over the base model across five financial and office‑task benchmarks.

By Hengqi Liu, Ziyuan Liu, Yang Qin, Wanting Zhou, Zichuan Wang, Jiachen Liang, Xu Chu, Shaowei Chen, Zhaokai Luo, Mu Chuan