SWE-Explore is a new benchmark evaluating repository exploration by coding agents on 848 issues across 203 repositories, using line-level ground truth from successful agent trajectories and showing agentic methods outperform classical retrieval on coverage and ranking.
hub Mixed citations
Swe-bench goes live!
Mixed citation behavior. Most common role is background (60%).
22
Pith papers citing it
1
external citations · Pith
Background
60% of classified citations
abstract
The issue-resolving task, where a model generates patches to fix real-world bugs, has emerged as a critical benchmark for evaluating the capabilities of large language models (LLMs). While SWE-bench and its variants have become standard in this domain, they suffer from key limitations: they have not been updated since their initial releases, cover a narrow set of repositories, and depend heavily on manual effort for instance construction and environment setup. These factors hinder scalability and introduce risks of overfitting and data contamination. In this work, we present SWE-bench-Live, a live-updatable benchmark designed to overcome these challenges. Our initial release consists of 1,319 tasks derived from real GitHub issues created since 2024, spanning 93 repositories. Each task is accompanied by a dedicated Docker image to ensure reproducible execution. Central to our benchmark is \method, an automated curation pipeline that streamlines the entire process from instance creation to environment setup, removing manual bottlenecks and enabling scalability and continuous updates. We evaluate a range of state-of-the-art agent frameworks and LLMs on SWE-bench-Live, revealing a substantial performance gap compared to static benchmarks like SWE-bench, even under controlled evaluation conditions. To better understand this discrepancy, we perform detailed analyses across repository origin, issue recency, and task difficulty. By providing a fresh, diverse, and executable benchmark grounded in live repository activity, SWE-bench-Live facilitates rigorous, contamination-resistant evaluation of LLMs and agents in dynamic, real-world software development settings.
hub tools
citation-role summary
background 2
dataset 1
method 1
other 1
citation-polarity summary
representative citing papers
RepoMirage uses semantics-preserving perturbations on SWE-Bench to show code agents lack repository context reasoning, with performance falling sharply on extended structure tasks, and introduces RepoAnchor as a structure-first fix.
AgentLens reveals 10.7% of passing SWE-agent trajectories exhibit Lucky Pass behaviors and introduces a process-level evaluation framework with a new annotated dataset of 1,815 trajectories.
SWE-WebDevBench finds that AI app builders commonly fail at translating business needs into complete, secure, production-ready software due to specification bottlenecks, frontend-backend decoupling, low engineering quality, and security weaknesses.
PlayCoder raises the rate of LLM-generated GUI apps that can be played end-to-end without logic errors from near zero to 20.3% Play@3 by adding repository-aware generation, agent-driven testing, and iterative repair.
Debug2Fix integrates interactive debugging via subagents into coding agents, delivering >20% gains on GitBug-Java and SWE-Bench-Live while enabling weaker models to match stronger ones.
SWE-EVO shows GPT-5.4 with OpenHands reaching only 25% success on complex multi-file evolution tasks versus 72.8% on SWE-Bench Verified, and introduces Fix Rate as a partial-progress metric.
Build-bench is the first architecture-aware benchmark that evaluates LLMs on repairing cross-ISA build failures via iterative tool-augmented reasoning, with the best model reaching 63.19% success.
Original, never-upstreamed multi-file engineering tasks with functional verifiers grade coding agents more faithfully and separate frontier models more widely than inherited-test SWE benchmarks.
An agentic code reviewer that explores repositories to judge and diagnose AI-generated pull requests improves resolve rates from 27.5% to 56.9% and outperforms single-turn review baselines.
SWE-Router introduces trajectory-conditioned value-based routing for LLM agents on SWE tasks, with a Bayes-optimality theorem and empirical cost savings while retaining most strong-model performance.
PROPEL amortizes solver evaluation with a trained activation probe to optimize task generators toward a target solve rate, raising the share of learnable tasks from ~10% to ~20% in coding and SWE experiments.
P2T distills reference patches into a latent process graph and uses it to select shortest effective trajectory segments from teacher rollouts, yielding up to 10.8 point Pass@1 gains on SWE-bench Verified with 15% lower inference cost using only 1.8k instances.
PITMuS reconstructs source-level original/mutated method pairs from PIT bytecode mutation reports at 99.96% coverage across eight Java projects.
SWE-Cycle benchmark shows sharp drops in code agent success rates from isolated tasks to full autonomous issue resolution, highlighting cross-phase dependency issues.
SkillSynth uses a scenario-mediated skill graph to sample workflow paths and generate executable terminal tasks, enabling controlled diversity in training trajectories for agents.
DryRUN lets LLMs create their own test inputs and run internal simulations for self-correcting code generation, matching the performance of test-dependent methods like CodeSIM on LiveCodeBench without public tests or external signals.
PlayCoder combines a repository-aware coding agent with a vision-based GUI testing agent and an automated program repair loop to detect and fix silent logic errors in LLM-generated interactive application code.
SWE-Bench Pro is a new benchmark with 1,865 long-horizon tasks from 41 repositories designed to evaluate AI agents on realistic enterprise-level software engineering problems beyond prior benchmarks.
icat-agent improves resolution rates on SWE-bench Verified and Pro by 3.6-18.5% over baselines via event-based multi-agent scaffolding and rubric-driven workflow pivoting while using the same models.
GLM-5 is a foundation model that claims state-of-the-art results on coding benchmarks and superior performance on end-to-end software engineering tasks via new asynchronous RL methods and cost-saving DSA.
EVT improves the RMT backbone by using Euclidean-distance attention decay and 1D token grouping, achieving 86.6% top-1 on ImageNet-1K at 384×384 resolution.
citing papers explorer
-
SWE-Explore: Benchmarking How Coding Agents Explore Repositories
SWE-Explore is a new benchmark evaluating repository exploration by coding agents on 848 issues across 203 repositories, using line-level ground truth from successful agent trajectories and showing agentic methods outperform classical retrieval on coverage and ranking.
-
RepoMirage: Probing Repository Context Reasoning in Code Agents with Perturbations
RepoMirage uses semantics-preserving perturbations on SWE-Bench to show code agents lack repository context reasoning, with performance falling sharply on extended structure tasks, and introduces RepoAnchor as a structure-first fix.
-
AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation
AgentLens reveals 10.7% of passing SWE-agent trajectories exhibit Lucky Pass behaviors and introduces a process-level evaluation framework with a new annotated dataset of 1,815 trajectories.
-
SWE-WebDevBench: Evaluating Coding Agent Application Platforms as Virtual Software Agencies
SWE-WebDevBench finds that AI app builders commonly fail at translating business needs into complete, secure, production-ready software due to specification bottlenecks, frontend-backend decoupling, low engineering quality, and security weaknesses.
-
PlayCoder: Making LLM-Generated GUI Code Playable
PlayCoder raises the rate of LLM-generated GUI apps that can be played end-to-end without logic errors from near zero to 20.3% Play@3 by adding repository-aware generation, agent-driven testing, and iterative repair.
-
Debug2Fix: Can Interactive Debugging Help Coding Agents Fix More Bugs?
Debug2Fix integrates interactive debugging via subagents into coding agents, delivering >20% gains on GitBug-Java and SWE-Bench-Live while enabling weaker models to match stronger ones.
-
SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarios
SWE-EVO shows GPT-5.4 with OpenHands reaching only 25% success on complex multi-file evolution tasks versus 72.8% on SWE-Bench Verified, and introduces Fix Rate as a partial-progress metric.
-
Can Language Models Go Beyond Coding? Assessing the Capability of Language Models to Build Real-World Systems
Build-bench is the first architecture-aware benchmark that evaluates LLMs on repairing cross-ISA build failures via iterative tool-augmented reasoning, with the best model reaching 63.19% success.
-
DeepSWE: Measuring Frontier Coding Agents on Original, Long-Horizon Engineering Tasks
Original, never-upstreamed multi-file engineering tasks with functional verifiers grade coding agents more faithfully and separate frontier models more widely than inherited-test SWE benchmarks.
-
SWE-Review: Closing the Loop on Issue Resolution with Agentic Code Review
An agentic code reviewer that explores repositories to judge and diagnose AI-generated pull requests improves resolve rates from 27.5% to 56.9% and outperforms single-turn review baselines.
-
SWE-Router: Routing in Multi-turn Agentic Software Engineering Tasks
SWE-Router introduces trajectory-conditioned value-based routing for LLM agents on SWE tasks, with a Bayes-optimality theorem and empirical cost savings while retaining most strong-model performance.
-
Breaking the Solver Bottleneck: Training Task Generators at the Learnable Frontier
PROPEL amortizes solver evaluation with a trained activation probe to optimize task generators toward a target solve rate, raising the share of learnable tasks from ~10% to ~20% in coding and SWE experiments.
-
From Patches to Trajectories: Privileged Process Supervision for Software-Engineering Agents
P2T distills reference patches into a latent process graph and uses it to select shortest effective trajectory segments from teacher rollouts, yielding up to 10.8 point Pass@1 gains on SWE-bench Verified with 15% lower inference cost using only 1.8k instances.
-
PITMuS: A Tool for Automated Bug Dataset Generation via Source-Level Mutant Reconstruction
PITMuS reconstructs source-level original/mutated method pairs from PIT bytecode mutation reports at 99.96% coverage across eight Java projects.
-
SWE-Cycle: Benchmarking Code Agents across the Complete Issue Resolution Cycle
SWE-Cycle benchmark shows sharp drops in code agent success rates from isolated tasks to full autonomous issue resolution, highlighting cross-phase dependency issues.
-
Toward Scalable Terminal Task Synthesis via Skill Graphs
SkillSynth uses a scenario-mediated skill graph to sample workflow paths and generate executable terminal tasks, enabling controlled diversity in training trajectories for agents.
-
You Don't Need Public Tests to Generate Correct Code
DryRUN lets LLMs create their own test inputs and run internal simulations for self-correcting code generation, matching the performance of test-dependent methods like CodeSIM on LiveCodeBench without public tests or external signals.
-
CityRAG: Stepping Into a City via Spatially-Grounded Video Generation
PlayCoder combines a repository-aware coding agent with a vision-based GUI testing agent and an automated program repair loop to detect and fix silent logic errors in LLM-generated interactive application code.
-
SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?
SWE-Bench Pro is a new benchmark with 1,865 long-horizon tasks from 41 repositories designed to evaluate AI agents on realistic enterprise-level software engineering problems beyond prior benchmarks.
-
Unlocking Model Potentials Through Adaptive Multi-Agent Scaffolding for Efficient Issue Resolution
icat-agent improves resolution rates on SWE-bench Verified and Pro by 3.6-18.5% over baselines via event-based multi-agent scaffolding and rubric-driven workflow pivoting while using the same models.
-
GLM-5: from Vibe Coding to Agentic Engineering
GLM-5 is a foundation model that claims state-of-the-art results on coding benchmarks and superior performance on end-to-end software engineering tasks via new asynchronous RL methods and cost-saving DSA.
-
ClawEnvKit: Automatic Environment Generation for Claw-Like Agents
EVT improves the RMT backbone by using Euclidean-distance attention decay and 1D token grouping, achieving 86.6% top-1 on ImageNet-1K at 384×384 resolution.