AI Paper Atlas

Research map / Large language models / Agents, code and math

Ekka: Automated Diagnosis of Silent Errors in LLM Inference

Yile Gu, Zhen Zhang, Shaowei Zhu, Xinwei Fu, Jun Wu, Yida Wang, Baris Kasikci

ICML 2026 · Large language models › Agents, code and math · approach: Code generation benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add mapping validation with error feedback to component mapping (ablation)Ekka component mapping · 9-bug subsetmapping accuracy: +21.7%Improved
add class-definition retrieval tool to component mapping (ablation)Ekka component mapping · 9-bug subsetmapping accuracy: +7.6%Improved
add codebase analysis tool and incremental mapping (ablation)Ekka · 9-bug subsetpass@1/pass@5: 0.67/0.77 -> 0.84/0.88Improved
add mapping validation/refinement loop (ablation)Ekka · 9-bug subsetpass@1/pass@5: 0.47/0.66 -> 0.84/0.88Improved
add code validation and error feedback to activation alignment (ablation)Ekka activation alignment · 9-bug subsetalignment accuracy: +54.9%Improved
add helper tools and knowledge base to activation alignment (ablation)Ekka activation alignment · 9-bug subsetalignment accuracy: +26%Improved
add helper tools and knowledge base for alignment (ablation)Ekka · 9-bug subset, ground-truth mappingpass@1/pass@5: 0.39/0.55 -> 0.84/0.88Improved
add alignment validation/refinement loop (ablation)Ekka · 9-bug subset, ground-truth mappingpass@1/pass@5: 0.69/0.77 -> 0.84/0.88Improved
raise error-ratio threshold from 1.5 to 2.0 or 2.5 (ablation)Ekka error analysis · bug benchmarkpass@1/pass@5: pass@1 0.84 -> 0.88 at 2.5; pass@5 0.88 unchangedNo clear effect
use Ekka differential debugging instead of general coding agentsEkka vs Mini-SWE-Agent and OpenCode · silent error benchmark, pass@1diagnosis accuracy improvement: +28% over Mini-SWE-Agent, +34% over OpenCodeImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy