Efficient SWE Agent Benchmarking via Trajectory-Aware Evaluation

• URL: http://arxiv.org/abs/2609.01603v1
• Date: Tue, 01 Sep 2026 17:59:46 GMT
• Title: Efficient SWE Agent Benchmarking via Trajectory-Aware Evaluation
• Abstract summary: We propose PTA-IRT, a Privileged Trajectory-Aware Item Response Theory framework that fuses process and outcome signals.<n>Under low calibration budgets, PTA-IRT consistently outperforms prior IRT baselines on score and ranking recovery across four SWE benchmarks.
• Score: 46.74484954631231
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Evaluating software engineering agents on realistic benchmarks is costly, since each task may require multi-step code exploration, modification, and test execution. Existing efficient evaluation methods select representative subsets to estimate full-benchmark performance, but are largely result-only: they fit historical pass/fail response matrices or static task semantics, discarding how agents solve problems. We propose PTA-IRT, a Privileged Trajectory-Aware Item Response Theory framework that fuses process and outcome signals. Historical execution trajectories supply process-level evidence beyond pass/fail, such as explored context, attempted edits, and solving paths, which PTA-IRT uses as privileged information for calibration subset selection and ability estimation. Under low calibration budgets, PTA-IRT consistently outperforms prior IRT baselines on score and ranking recovery across four SWE benchmarks. Code and data are publicly available at https://github.com/DeepSoftwareAnalytics/PTA-IRT.

Related papers

• PERFOPT-Bench: Evaluating Coding Agents on Software Performance Optimization [18.924749191192433]
  Performance optimization is a distinct agentic task.<n>We introduce PERFOPT-Bench, a benchmark for evaluating this full performance-engineering loop.
  arXiv  Detail & Related papers  (2026-07-08T08:32:18Z)
• Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis [68.28714988482703]
  Process Reward Models (PRMs) have achieved remarkable success in augmenting the reasoning capabilities of Large Language Models (LLMs)<n>We present a study revealing that general-domain PRMs struggle to supervise data analysis agents.<n>We introduce DataPRM, a novel environment-aware generative process reward model.
  arXiv  Detail & Related papers  (2026-04-27T09:00:30Z)
• SWE-TRACE: Optimizing Long-Horizon SWE Agents Through Rubric Process Reward Models and Heuristic Test-Time Scaling [13.335296846555204]
  Current pipelines are bottlenecked by unoptimized demonstration data, sparse execution rewards, and computationally prohibitive inference scaling.<n>We present SWE-TRACE, a unified framework optimizing the SWE agent lifecycle across data curation, reinforcement learning (RL), and test-time inference.<n> experiments on standard SWE benchmarks demonstrate that SWE-TRACE significantly advances the state-of-the-art, maximizing resolution rates while drastically reducing both token consumption inference latency.
  arXiv  Detail & Related papers  (2026-04-16T09:41:47Z)
• Runtime Execution Traces Guided Automated Program Repair with Multi-Agent Debate [8.424102114588559]
  Automated Program Repair (APR) struggles with complex logic errors and silent failures.<n>Current LLM-based APR methods are mostly static, relying on source code and basic test outputs.<n>We propose TraceRepair, a multi-agent framework that leverages runtime facts as shared constraints for patch validation.
  arXiv  Detail & Related papers  (2026-04-03T02:23:25Z)
• ELT-Bench-Verified: Benchmark Quality Issues Underestimate AI Agent Capabilities [4.5258165293324515]
  Extract-Load-Transform pipelines are a labor-intensive data engineering task and a high-impact target for AI automation.<n>On ELT-Bench, the first benchmark for end-to-end ELT pipeline construction, AI agents initially showed low success rates.<n>We revisit these results and identify two factors causing a substantial underestimation of agent capabilities.
  arXiv  Detail & Related papers  (2026-03-31T08:02:16Z)
• Verified Critical Step Optimization for LLM Agents [67.05296684575445]
  Critical Step Optimization focuses preference learning on verified critical steps.<n>Method starts from failed policy trajectories rather than expert demonstrations.<n>Experiments on GAIA-Text-103 and XBench-DeepSearch show that CSO achieves 37% and 26% relative improvement over the SFT baseline.
  arXiv  Detail & Related papers  (2026-02-03T11:41:02Z)
• Guided by Trajectories: Repairing and Rewarding Tool-Use Trajectories for Tool-Integrated Reasoning [65.10602992874787]
  AutoTraj is a framework that automatically learns TIR by repairing and rewarding tool-use trajectories.<n> Experiments on real-world benchmarks demonstrate the effectiveness of AutoTraj.
  arXiv  Detail & Related papers  (2026-01-30T14:42:04Z)
• BOTS: A Unified Framework for Bayesian Online Task Selection in LLM Reinforcement Finetuning [82.925106913459]
  Reinforcement finetuning (RFT) is a key technique for aligning Large Language Models (LLMs) with human preferences and enhancing reasoning.<n>We introduce BOTS, a unified framework for Bayesian Online Task Selection in RFT reinforcement finetuning.
  arXiv  Detail & Related papers  (2025-10-30T11:15:23Z)
• Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents [22.781523439717223]
  A proper evaluation of an agent's performance must go beyond the final answer to also assess the problem-solving trajectory.<n>We introduce TRACE, a framework for the multi-dimensional evaluation of tool-augmented LLM agent performance.<n>Our results confirm that TRACE accurately evaluates these complex behaviors in a scalable and cost-effective manner.
  arXiv  Detail & Related papers  (2025-10-03T09:19:15Z)
• Re-Evaluating LiDAR Scene Flow for Autonomous Driving [80.37947791534985]
  Popular benchmarks for self-supervised LiDAR scene flow have unrealistic rates of dynamic motion, unrealistic correspondences, and unrealistic sampling patterns. We evaluate a suite of top methods on a suite of real-world datasets. We show that despite the emphasis placed on learning, most performance gains are caused by pre- and post-processing steps.
  arXiv  Detail & Related papers  (2023-04-04T22:45:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.