RuBench, the first repository-level coding benchmark with natively authored (non-translated) Russian task specifications, measures deployed coding agents on 25 contamination-gated fix tasks and documents model substitution and answer leakage that change how agent benchmarks must be audited.
hub Canonical reference
SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?
Canonical reference. 76% of citing Pith papers cite this work as background.
95
Pith papers citing it
Background
76% of classified citations
abstract
We introduce SWE-Bench Pro, a substantially more challenging benchmark that builds upon the best practices of SWE-BENCH [25], but is explicitly designed to capture realistic, complex, enterprise-level problems beyond the scope of SWE-BENCH. SWE-BENCH PRO contains 1,865 problems sourced from a diverse set of 41 actively maintained repositories spanning business applications, B2B services, and developer tools. The benchmark is partitioned into a public set with open access to problems sourced from 11 repositories, a held-out set of 12 repositories and a commercial set of 18 proprietary repositories where we have formal partnership agreements with early-stage startups. Problems in the held-out and the commercial set are not publicly accessible, but we release results on the commercial set. Our benchmark features long-horizon tasks that may require hours to days for a professional software engineer to complete, often involving patches across multiple files and substantial code modifications. All tasks are human-verified and augmented with sufficient context to ensure resolvability. To better understand these limitations, we cluster the failure modes observed in the collected agent trajectories for a clearer characterization of the error patterns exhibited by current models. Overall, SWE-BENCH PRO provides a contamination-resistant testbed that more faithfully captures the complexity and diversity of real-world software development, advancing the pursuit of truly autonomous software engineering agents at a professional level.
hub tools
citation-role summary
background 10
dataset 6
baseline 1
citation-polarity summary
representative citing papers
CL-Bench is the first expert-validated benchmark for continual learning in frontier LLMs across six real-world domains, showing limited gains and that naive in-context learning outperforms dedicated memory systems.
PDEAgent-Bench is the first multi-metric, multi-library benchmark for AI-generated PDE solvers, evaluating executability, numerical accuracy, and efficiency across DOLFINx, Firedrake, and deal.II.
HWE-Bench is the first repository-level benchmark for LLM agents on real hardware bug repair, where the best agent fixes 70.7% of 417 tasks but drops below 65% on complex SoC projects.
SWE-Interact shows frontier models solve roughly 25% of multi-turn interactive coding tasks versus 50% on single-turn baselines.
On 108 long-horizon real-world computer workflows, frontier agents complete at most 20.6% of tasks and fail mainly by losing hidden state, not by basic GUI control.
Dockerless uses agentic repository exploration to verify patches without execution, enabling SFT and RL training of coding agents that reach 62.0/50.0/35.2% resolve rates on SWE-bench Verified/Multilingual/Pro while matching environment-based results.
CFAgentBench is a new reproducible benchmark for construction-finance AI agents featuring 35 mock apps, 1,014 tasks, and a money-movement guard, with initial tests showing pass^1 of 0.67 dropping to pass^5 of 0.38.
StaminaBench evaluates coding agents over 100 procedurally generated change requests to a REST API, finding that tested models fail within 5-6 turns without feedback but improve up to 12x with test feedback and good harnesses.
MAFP applies fictitious play to LLM multi-agent systems to resolve stance entanglement in competitive decision-making, outperforming single-round and multi-round baselines on tournament strength and robustness.
AgentBeats implements agentified evaluation of diverse AI agents through standardized interfaces, validated at scale in a five-month competition with 298 judges and 467 subjects plus a coding case study.
Claw-SWE-Bench is a 350-instance multilingual benchmark for OpenClaw-style agent harnesses that shows adapter design raises Pass@1 from 19.1% to 73.4% on the same model while releasing data for reproducible comparison.
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.
TensorBench is a new benchmark of 199 tasks on a tensor framework used to evaluate seven coding agents, yielding pass rates from 22.1% to 64.8% with low inter-agent agreement.
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.
SpecBench shows frontier coding agents saturate visible test suites but exhibit persistent reward hacking on held-out tests, with the gap growing 28 percentage points per tenfold increase in code size.
SaaSBench introduces a heterogeneous benchmark for enterprise SaaS engineering and shows that state-of-the-art coding agents fail over 95% of the time before reaching deep business logic due to setup and integration problems.
BenchJack audits 10 AI agent benchmarks, synthesizes exploits achieving near-perfect scores without task completion, surfaces 219 flaws, and reduces hackable-task ratios to under 10% on four benchmarks via iterative patching.
LLM agents encode tool necessity in pre-generation hidden states with high linear decodability (AUROC 0.89-0.96); Probe&Prefill uses this to reduce tool calls 48% with 1.7% accuracy loss.
Goal clarifications lose nearly all value after 10% of execution while input clarifications retain value until roughly 50%, and asking any type past mid-trajectory hurts performance more than never asking.
LLM agents exhibit constraint decay with assertion pass rates dropping substantially as structural requirements increase in multi-file backend code generation across web frameworks.
TEBench is a new project-level benchmark for test evolution showing coding agents achieve only 45-49% F1 on identifying tests needing changes, with stale tests hardest due to reliance on execution failures.
ProgramBench introduces 200 tasks where models must reconstruct full programs like FFmpeg or SQLite from docs alone; none of 9 evaluated LMs fully solve any task and the best passes 95% tests on only 3% of tasks while favoring monolithic code.
SkillFlow benchmark shows lifelong skill evolution yields modest gains for some models like Claude Opus 4.6 but limited or negative utility for others despite high skill usage.
citing papers explorer
-
RuBench: A Repository-Level Agentic Coding Benchmark with Natively Authored Russian Task Specifications
RuBench, the first repository-level coding benchmark with natively authored (non-translated) Russian task specifications, measures deployed coding agents on 25 contamination-gated fix tasks and documents model substitution and answer leakage that change how agent benchmarks must be audited.
-
Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments
CL-Bench is the first expert-validated benchmark for continual learning in frontier LLMs across six real-world domains, showing limited gains and that naive in-context learning outperforms dedicated memory systems.
-
PDEAgent-Bench: A Multi-Metric, Multi-Library Benchmark for PDE Solver Generation
PDEAgent-Bench is the first multi-metric, multi-library benchmark for AI-generated PDE solvers, evaluating executability, numerical accuracy, and efficiency across DOLFINx, Firedrake, and deal.II.
-
HWE-Bench: Benchmarking LLM Agents on Real-World Hardware Bug Repair Tasks
HWE-Bench is the first repository-level benchmark for LLM agents on real hardware bug repair, where the best agent fixes 70.7% of 417 tasks but drops below 65% on complex SoC projects.
-
SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions
SWE-Interact shows frontier models solve roughly 25% of multi-turn interactive coding tasks versus 50% on single-turn baselines.
-
OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks
On 108 long-horizon real-world computer workflows, frontier agents complete at most 20.6% of tasks and fail mainly by losing hidden state, not by basic GUI control.
-
Dockerless: Environment-Free Program Verifier for Coding Agents
Dockerless uses agentic repository exploration to verify patches without execution, enabling SFT and RL training of coding agents that reach 62.0/50.0/35.2% resolve rates on SWE-bench Verified/Multilingual/Pro while matching environment-based results.
-
CFAgentBench: A Reproducible Environment and Benchmark for Autonomous Construction-Finance Agents
CFAgentBench is a new reproducible benchmark for construction-finance AI agents featuring 35 mock apps, 1,014 tasks, and a money-movement guard, with initial tests showing pass^1 of 0.67 dropping to pass^5 of 0.38.
-
StaminaBench: Stress-Testing Coding Agents over 100 Interaction Turns
StaminaBench evaluates coding agents over 100 procedurally generated change requests to a REST API, finding that tested models fail within 5-6 turns without feedback but improve up to 12x with test feedback and good harnesses.
-
Enhancing Decision-Making with Large Language Models through Multi-Agent Fictitious Play
MAFP applies fictitious play to LLM multi-agent systems to resolve stance entanglement in competitive decision-making, outperforming single-round and multi-round baselines on tournament strength and robustness.
-
AgentBeats: Agentifying Agent Assessment for Openness, Standardization, and Reproducibility
AgentBeats implements agentified evaluation of diverse AI agents through standardized interfaces, validated at scale in a five-month competition with 298 judges and 467 subjects plus a coding case study.
-
Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding Tasks
Claw-SWE-Bench is a 350-instance multilingual benchmark for OpenClaw-style agent harnesses that shows adapter design raises Pass@1 from 19.1% to 73.4% on the same model while releasing data for reproducible comparison.
-
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.
-
TensorBench: Benchmarking Coding Agents on a Compiler-Based Tensor Framework
TensorBench is a new benchmark of 199 tasks on a tensor framework used to evaluate seven coding agents, yielding pass rates from 22.1% to 64.8% with low inter-agent agreement.
-
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.
-
SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agents
SpecBench shows frontier coding agents saturate visible test suites but exhibit persistent reward hacking on held-out tests, with the gap growing 28 percentage points per tenfold increase in code size.
-
SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering
SaaSBench introduces a heterogeneous benchmark for enterprise SaaS engineering and shows that state-of-the-art coding agents fail over 95% of the time before reaching deep business logic due to setup and integration problems.
-
Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack
BenchJack audits 10 AI agent benchmarks, synthesizes exploits achieving near-perfect scores without task completion, surfaces 219 flaws, and reduces hackable-task ratios to under 10% on four benchmarks via iterative patching.
-
LLM Agents Already Know When to Call Tools -- Even Without Reasoning
LLM agents encode tool necessity in pre-generation hidden states with high linear decodability (AUROC 0.89-0.96); Probe&Prefill uses this to reduce tool calls 48% with 1.7% accuracy loss.
-
Ask Early, Ask Late, Ask Right: When Does Clarification Timing Matter for Long-Horizon Agents?
Goal clarifications lose nearly all value after 10% of execution while input clarifications retain value until roughly 50%, and asking any type past mid-trajectory hurts performance more than never asking.
-
Constraint Decay: The Fragility of LLM Agents in Backend Code Generation
LLM agents exhibit constraint decay with assertion pass rates dropping substantially as structural requirements increase in multi-file backend code generation across web frameworks.
-
Breaking, Stale, or Missing? Benchmarking Coding Agents on Project-Level Test Evolution
TEBench is a new project-level benchmark for test evolution showing coding agents achieve only 45-49% F1 on identifying tests needing changes, with stale tests hardest due to reliance on execution failures.
-
ProgramBench: Can Language Models Rebuild Programs From Scratch?
ProgramBench introduces 200 tasks where models must reconstruct full programs like FFmpeg or SQLite from docs alone; none of 9 evaluated LMs fully solve any task and the best passes 95% tests on only 3% of tasks while favoring monolithic code.
-
SkillFlow:Benchmarking Lifelong Skill Discovery and Evolution for Autonomous Agents
SkillFlow benchmark shows lifelong skill evolution yields modest gains for some models like Claude Opus 4.6 but limited or negative utility for others despite high skill usage.
-
Problem Reductions at Scale: Agentic Integration of Computationally Hard Problems
A harness for AI agents enabled construction of a Rust library with 100+ problem types and 200+ reduction rules for NP-hard problems in three months.
-
HiL-Bench (Human-in-Loop Benchmark): Do Agents Know When to Ask for Help?
HiL-Bench shows frontier AI agents fail to ask for help on incomplete tasks, recovering only a fraction of full-information performance, but RL training on Ask-F1 reward improves judgment and transfers across domains.
-
Inside the Scaffold: A Source-Code Taxonomy of Coding Agent Architectures
Analysis of 13 coding agent scaffolds at pinned commits yields a 12-dimension taxonomy showing five composable loop primitives, with 11 agents combining multiple primitives instead of using one fixed structure.
-
AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents
AgentHazard benchmark shows computer-use agents remain highly vulnerable, with attack success rates reaching 73.63% on models like Qwen3-Coder powering Claude Code.
-
Beyond Resolution Rates: Behavioral Drivers of Coding Agent Success and Failure
Large-scale trajectory analysis of 19 coding agents on 500 tasks finds that LLM choice drives outcomes more than framework design and that context-gathering plus validation behaviors improve success beyond task difficulty predictions.
-
Vibe Code Bench: Evaluating AI Models on End-to-End Web Application Development
Vibe Code Bench evaluates AI models on building complete web applications from specs, with the best of 16 models achieving 61.8% accuracy on the test split using autonomous browser evaluation.
-
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.
-
Agent-Diff: Benchmarking LLM Agents on Enterprise API Tasks via Code Execution with State-Diff-Based Evaluation
Agent-Diff benchmarks LLM agents on enterprise API tasks using code execution and state-diff contracts to define success, evaluated on nine models across 224 tasks with code released.
-
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.
-
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.
-
What Makes a Good Bug Report for an AI Agent?
AI repair agents solve bugs more reliably when reports include executable reproduction scripts, file-level localization cues, and clear structure, while longer prose reports and human-oriented steps to reproduce show no benefit or hurt.
-
What Resolve Rate Hides: Trajectory Structure Diagnostics for Coding Agents
TraceProbe normalizes coding agent trajectories into canonical actions and applies rule-based detectors to localize failure patterns and behavioral divergences that resolve rate hides.
-
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.
-
Beyond Textual Repository Exploration: Dual-Modal Structural Reasoning for Agentic Issue Resolution
DUALVIEW is a dual-modal framework using Module Coupling, Function Call, Class Hierarchy, and Program Dependence graphs to enable persistent structural reasoning for agentic issue resolution, reporting gains on SWE-bench Pro and Verified.
-
Cheap Code, Costly Judgment: A Case Study on Governable Agentic Software Engineering
High-velocity agentic coding becomes governable when engineers convert recurring structural failures into durable, machine-actionable governance mechanisms rather than relying on continuous human code review.
-
LLVM-Bench: Benchmarking and Advancing Large Language Models for LLVM Compiler Issue Resolution
LLVM-Bench supplies 423 validated LLVM issues and LLVM-Gym automates evaluation, showing LLMs are limited but an ensemble reaches 21.99% resolution.
-
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.
-
MirrorCode: AI can rebuild entire programs from behavior alone
AI agents given only executable behavior and tests can reimplement fully scoped programs (e.g., gotree: 16k LoC, 2000/2001 tests), and the best model solves 56% of 25 MirrorCode tasks.
-
Life After Benchmark Saturation: A Case Study of CORE-Bench
Using CORE-Bench as a case study, the paper shows that saturated benchmarks can still deliver insights on efficiency, reliability, model-scaffold differences, and human collaboration even after accuracy plateaus, and introduces improved benchmark versions plus a small randomized experiment demonstra
-
Qwen-AgentWorld: Language World Models for General Agents
Qwen-AgentWorld are language world models that simulate multi-domain agent environments and boost general agent capabilities via decoupled RL simulation and unified foundation model training.
-
Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation
AFTER benchmark shows single refinement improves LLM agent performance by 3.7-6.7 points and multi-model procedural skills reach 73.1% cross-model accuracy on 382 tasks.
-
A Framework for Evaluating Agentic Skills at Scale
The authors developed an evaluation framework that generates 1000 tasks from 500 real-world agent skills, applies instruction-following and goal-completion rubrics, and benchmarks 19 proprietary and open-source model configurations.
-
Position: Coding Benchmarks Are Misaligned with Agentic Software Engineering
Coding benchmarks conflate the model with the harness, environment, and verifier into a single end-to-end score, which is misaligned with agentic software engineering.
-
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.
-
DeskCraft: Benchmarking Desktop Agents on Professional Workflows and Human-in-the-Loop Collaboration
DeskCraft provides 538 tasks across design, video, audio, and 3D software with a multilevel taxonomy and formalized mid-turn and post-turn human-agent interaction protocols, evaluating 18 agents with top performance at 31.6% on standard tasks.
-
Idleness is Relative: Exploiting Tool-Call Idle Windows for Offloading in Agentic Systems with MORI
MORI improves throughput 20-71% and TTFT 18-43% over baselines by ranking programs on a continuous idleness spectrum and shifting the GPU-CPU boundary to match capacity in agentic LLM serving.