Small coding agents gain little from code graphs past one hop
On SWE-bench Verified, structured tools added 17–21 points at 1.7–4B but only 4–5 points at 32B.
46
Product
SWE-bench Verified
- Mentions · 7d0
- Mentions · 30d5
- All mentions9
- First seenSep 16, 2026
- Last mentioned8d ago
Timeline
9 articles · newest firstIndependent review lifts coding-agent accuracy at modest extra cost
A critic review gains 5.9–9.5 points on Terminal-Bench and 3.9–6.6 on SWE-bench Verified at about baseline cost.
48
Show HN: Parsimony – an open-source benchmark for how much code AI models add
Parsimony ranks AI coding models by net syntax units added on SWE-bench Verified patches.
34
ANTE: Anticipate, Then Execute -- Learning Internal World Models for Coding Agents from Verified Beliefs
ANTE lifts Qwen3-32B to 57.2% pass@1 on SWE-bench Verified using verified action forecasts.
58
FC-SWE trains coding agents on recovery after failed patches
With Qwen3.5-4B and SWE-agent it reports 41.7% Resolved@1 versus 38.9% for GRPO.
50
PivotOPD: Learning to Recover from Pivotal Mistakes in Multi-Turn Agents
PivotOPD trains multi-turn agents to avoid pivotal mistakes and recover within a few turns.
52
An Empirical Study of Harness Design for Coding Agents
Empirical study of coding agent harnesses across 176 settings shows context management prevents overflow failures and bash-only interfaces cut costs for capable models.
35
An Empirical Study of Harness Design for Coding Agents
Empirical study of 176 matched coding-agent harness configurations shows context management cuts overflow failures while planning and action-space choices trade accuracy against cost.
42
Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
DualViewEval compresses agent benchmarks by jointly modeling outcome and process signals, achieving 24x-40x compression with only 20 tasks on APEX-Agents and BFCL.
30
Appears with
- Terminal-Bench 2.1product · 2
- Qwen3-8Bai model · 1
- VisualBenchproduct · 1
- Qwen3.5-4Bai model · 1
- APEX-Agentsproduct · 1
- Qwen3-32Bai model · 1
- Qwen3-1.7Bai model · 1
- Qwen3ai model · 1
- Gemma 4ai model · 1
- mini-SWE-agentproduct · 1
- NVIDIAorganization · 1
- Nemotron-3.5ai model · 1
- SWE-agentproduct · 1
- NVIDIAvendor · 1
- WebShopproduct · 1
- Parsimonyproduct · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.