{"harness": {"slug": "swe-fixer", "name": "SWE Fixer", "vendor": "Internlm", "description": "SWE Fixer is tracked in this catalog as a software-engineering agent harness. Public benchmark rows use this slug on leaderboards such as SWE-bench Verified or Artificial Analysis coding-agent evals.\n\nCheck each score row's `source` and `source_url` for the authoritative snapshot. Harness-level metadata here does not invent scores \u2014 only documents what the harness is and where published results come from.", "card_summary": "Internlm SWE Fixer coding agent harness.", "homepage_url": "https://github.com/InternLM/SWE-Fixer", "repo_url": "https://github.com/InternLM/SWE-Fixer", "docs_url": null, "scaffold_code": null, "harbor_agent_name": null, "api_protocol": null, "status": "validated", "supports_rl": true, "score_count": 4, "best_success_rate": 32.8, "github_stars": 139, "popularity_tier": null, "hrl_score": 29.0, "hrl_rank": 41, "catalog_token_volume": null, "openrouter_icon_url": null, "leaderboard_icon_url": "https://www.swebench.com/img/logos/20241128_SWE-Fixer_Qwen2.5-7b-retriever_Qwen2.5-72b-editor_20241128.png", "aa_coding_index": null, "aa_mean_cost_usd": null, "aa_mean_total_tokens": null}, "scores": [{"success_rate": 32.8, "resolved_count": 164, "total_count": 500, "cost_usd": null, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Verified public submission", "source": "swe_bench_site", "source_url": "https://github.com/InternLM/SWE-Fixer", "observed_at": "2025-03-06T00:00:00+00:00", "harness": {"slug": "swe-fixer", "name": "SWE Fixer", "vendor": "Internlm", "repo_url": "https://github.com/InternLM/SWE-Fixer", "openrouter_icon_url": null}, "model": {"slug": "ensemble-swe-fixer-qwen2-5-7b-retriever-qwen2-5-72b-editor", "name": "Ensemble (SWE Fixer (Qwen2.5 7b Retriever, Qwen2.5 72b Editor)", "vendor": "Qwen"}, "benchmark": {"slug": "swe-bench-verified", "name": "SWE-bench Verified"}}, {"success_rate": 30.2, "resolved_count": 151, "total_count": 500, "cost_usd": null, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Verified public submission", "source": "swe_bench_site", "source_url": "https://github.com/InternLM/SWE-Fixer", "observed_at": "2024-11-28T00:00:00+00:00", "harness": {"slug": "swe-fixer", "name": "SWE Fixer", "vendor": "Internlm", "repo_url": "https://github.com/InternLM/SWE-Fixer", "openrouter_icon_url": null}, "model": {"slug": "ensemble-swe-fixer-qwen2-5-7b-retriever-qwen2-5-72b-editor-20241128", "name": "Ensemble (SWE Fixer (Qwen2.5 7b Retriever, Qwen2.5 72b Editor) 20241128)", "vendor": "Qwen"}, "benchmark": {"slug": "swe-bench-verified", "name": "SWE-bench Verified"}}, {"success_rate": 24.67, "resolved_count": 74, "total_count": 300, "cost_usd": null, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Lite public submission", "source": "swe_bench_site", "source_url": "https://github.com/InternLM/SWE-Fixer", "observed_at": "2025-03-06T00:00:00+00:00", "harness": {"slug": "swe-fixer", "name": "SWE Fixer", "vendor": "Internlm", "repo_url": "https://github.com/InternLM/SWE-Fixer", "openrouter_icon_url": null}, "model": {"slug": "ensemble-swe-fixer-qwen2-5-7b-retriever-qwen2-5-72b-editor", "name": "Ensemble (SWE Fixer (Qwen2.5 7b Retriever, Qwen2.5 72b Editor)", "vendor": "Qwen"}, "benchmark": {"slug": "swe-bench-lite", "name": "SWE-bench Lite"}}, {"success_rate": 23.33, "resolved_count": 70, "total_count": 300, "cost_usd": null, "input_tokens": null, "output_tokens": null, "latency_ms": null, "temperature": null, "max_turns": null, "context_budget_tokens": null, "harness_version": null, "trial_count": null, "is_harnessrl_measured": false, "eval_protocol": "SWE-bench Lite public submission", "source": "swe_bench_site", "source_url": "https://github.com/InternLM/SWE-Fixer", "observed_at": "2024-11-28T00:00:00+00:00", "harness": {"slug": "swe-fixer", "name": "SWE Fixer", "vendor": "Internlm", "repo_url": "https://github.com/InternLM/SWE-Fixer", "openrouter_icon_url": null}, "model": {"slug": "ensemble-swe-fixer-qwen2-5-7b-retriever-qwen2-5-72b-editor-20241128", "name": "Ensemble (SWE Fixer (Qwen2.5 7b Retriever, Qwen2.5 72b Editor) 20241128)", "vendor": "Qwen"}, "benchmark": {"slug": "swe-bench-lite", "name": "SWE-bench Lite"}}], "benchmarks": [{"slug": "swe-bench-lite", "name": "SWE-bench Lite", "category": "eval"}, {"slug": "swe-bench-verified", "name": "SWE-bench Verified", "category": "eval"}]}