c-CRAB benchmark shows state-of-the-art code review agents solve only around 40% of tasks derived from human reviews, suggesting potential for human-AI collaboration.
Featurebench: Benchmarking agentic coding for complex feature development
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 8roles
background 1polarities
background 1representative citing papers
Coding-agent performance is workload- and framework-dependent, and raw speedup is an unsafe score because agents exploit benchmark-specific shortcuts.
Agent-as-a-Router turns static LLM routing into an iterative C-A-F loop that accumulates execution feedback to lower cumulative regret on coding tasks.
SWE-Cycle benchmark shows sharp drops in code agent success rates from isolated tasks to full autonomous issue resolution, highlighting cross-phase dependency issues.
ATM is a CID-brokered governance framework that maps write intents to semantic atoms for pre-admission control, validation, and neutral-steward application in single-domain multi-agent code synthesis.
Insights Generator is a multi-agent system that produces evidence-backed insights from corpora of LLM agent traces and yields 30.4pp performance gains when humans apply the reports.
Agentic Consensus replaces code as the main artifact with a typed property graph world model that maintains commitments and evidence through synchronization operators, shifting evaluation to alignment fidelity and consensus entropy.
Survey framing LLM agents as model-plus-harness systems, decomposing harness responsibilities, mapping them to tasks, and highlighting open challenges in evaluation, safety, and co-evolution.
citing papers explorer
-
Code Review Agent Benchmark
c-CRAB benchmark shows state-of-the-art code review agents solve only around 40% of tasks derived from human reviews, suggesting potential for human-AI collaboration.
-
PERFOPT-Bench: Evaluating Coding Agents on Software Performance Optimization
Coding-agent performance is workload- and framework-dependent, and raw speedup is an unsafe score because agents exploit benchmark-specific shortcuts.
-
Agent-as-a-Router: Agentic Model Routing for Coding Tasks
Agent-as-a-Router turns static LLM routing into an iterative C-A-F loop that accumulates execution feedback to lower cumulative regret on coding tasks.
-
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.
-
ATM: CID-Brokered Pre-Write Admission for Multi-Agent Code Co-Synthesis
ATM is a CID-brokered governance framework that maps write intents to semantic atoms for pre-admission control, validation, and neutral-steward application in single-domain multi-agent code synthesis.
-
Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents
Insights Generator is a multi-agent system that produces evidence-backed insights from corpora of LLM agent traces and yields 30.4pp performance gains when humans apply the reports.
-
Scaling Human-AI Coding Collaboration Requires a Governable Consensus Layer
Agentic Consensus replaces code as the main artifact with a typed property graph world model that maintains commitments and evidence through synchronization operators, shifting evaluation to alignment fidelity and consensus entropy.
-
From Question Answering to Task Completion: A Survey on Agent System and Harness Design
Survey framing LLM agents as model-plus-harness systems, decomposing harness responsibilities, mapping them to tasks, and highlighting open challenges in evaluation, safety, and co-evolution.