Measuring LLM Trust Allocation Across Conflicting Software Artifacts

• URL: http://arxiv.org/abs/2604.03447v1
• Date: Fri, 03 Apr 2026 20:38:14 GMT
• Title: Measuring LLM Trust Allocation Across Conflicting Software Artifacts
• Abstract summary: We present TRACE, a framework that elicits structured artifact-level trust traces over Javadoc, method signatures, implementations, and test prefixes.<n>We evaluate per-artifact quality assessment, inconsistency detection, affected artifact attribution, and source prioritization.
• Score: 4.974612243038232
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LLM-based software engineering assistants fail not only by producing incorrect outputs, but also by allocating trust to the wrong artifact when code, documentation, and tests disagree. Existing evaluations focus mainly on downstream outcomes and therefore cannot reveal whether a model recognized degraded evidence, identified the unreliable source, or calibrated its trust across artifacts. We present TRACE (Trust Reasoning over Artifacts for Calibrated Evaluation), a framework that elicits structured artifact-level trust traces over Javadoc, method signatures, implementations, and test prefixes under blind perturbations. Using 22,339 valid traces from seven models on 456 curated Java method bundles, we evaluate per-artifact quality assessment, inconsistency detection, affected artifact attribution, and source prioritization. Across all models, quality penalties are largely localized to the perturbed artifact and increase with severity, but sensitivity is asymmetric across artifact types: documentation bugs induce a substantially larger heavy-to-subtle gap than implementation faults (0.152-0.253 vs. 0.049-0.123). Models detect explicit documentation bugs well (67-94%) and Javadoc and implementation contradictions at 50-91%, yet show a systematic blind spot when only the implementation drifts while the documentation remains plausible, with detection dropping by 7-42 percentage points. Confidence is poorly calibrated for six of seven models. These findings suggest that current LLMs are better at auditing natural-language specifications than at detecting subtle code-level drift, motivating explicit artifact-level trust reasoning before correctness-critical downstream use.

Related papers

• IdeaAMBIG: Benchmarking Implementation-Critical Gaps in Research-Idea Specifications [52.570108663867046]
  A research idea may be novel, coherent, and scientifically plausible, yet its proposed method may remain insufficiently specified for faithful implementation.<n>We study the codification readiness of implementation-facing research-method specifications, defined by whether they provide sufficient methodological information for a competent implementer or coding agent to construct the intended method without unsupported assumptions.<n>We introduce IdeaAMBIG, a benchmark of 660 evidence-grounded instances: 163 real-world gaps from reports and GitHub issues, and 497 controlled synthetic gaps injected into codification-ready references.
  arXiv  Detail & Related papers  (2026-09-09T17:59:04Z)
• Diagnosing Fine-Grained Inconsistency Classification in Financial Disclosure Text [3.2712318205295907]
  Financial disclosures contain numerical claims, temporal statements, entity references, policy commitments, and risk descriptions that may conflict in qualitatively different ways.<n>We study this problem as fine-grained inconsistency.<n>A fine-tuned 300M encoder reaches 61.9% accuracy, compared with 61.5% for a LoRA-adapted Qwen3.5-9B model and 61.3% for GPT-adapted generative models.
  arXiv  Detail & Related papers  (2026-07-29T01:03:53Z)
• Faults in Our Formal Benchmarking: Dataset Defects and Evaluation Failures in Lean Theorem Proving [13.87786274932945]
  Benchmarks for LLM-assisted theorem proving in Lean are often treated as intrinsically reliable because every solved instance comes with a machine-checked proof.<n>We audit five widely used Lean theorem-proving benchmarks and their forks, using corpus-scale static checkers to surface 4,833 findings.<n>We propose a fault taxonomy, a suite of automated checkers and recall-oriented semantic audit prompts, and release standards to guide the creation of formal math datasets.
  arXiv  Detail & Related papers  (2026-06-28T16:46:36Z)
• SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation [82.22252244442744]
  SYNCRED-Bench is a benchmark of 600 AI-generated misinformation images balanced across six credible-form categories and seven fine-grained circulation styles.<n>Under a 5% false-positive-rate constraint, 15 MLLMs achieve only 10.5% true positive rate (TPR), open-source AIGC detectors achieve less than 5%.<n>Human annotators also struggled to identify synthetic credibility, reaching only 63% TPR.
  arXiv  Detail & Related papers  (2026-06-02T08:57:38Z)
• How Compliant Are GitHub Actions Workflows? A Checklist-Based Study with LLM-Assisted Auditing [0.0]
  GitHub Actions (GHA) CI are critical infrastructure, but current tooling offers only syntactic or checks.<n>This paper proposes a novel, documentation-grounded GHA compliance checklist with 30 criteria spanning four workflow sections and eight themes.
  arXiv  Detail & Related papers  (2026-05-03T23:21:13Z)
• HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam [63.84155758655084]
  Humanity's Last Exam (HLE) has become a widely used benchmark for evaluating frontier large language models.<n>We introduce HLE-Verified, a verified and revised version of HLE with a transparent verification protocol and fine-grained error taxonomy.<n>We evaluate seven state-of-the-art language models on HLE and HLE-Verified, observing an average absolute accuracy gain of 7--10 percentage points.
  arXiv  Detail & Related papers  (2026-02-15T02:50:15Z)
• CodeFuse-CommitEval: Towards Benchmarking LLM's Power on Commit Message and Code Change Inconsistency Detection [8.631593963090985]
  Version control relies on commit messages to convey the rationale for code changes, but these messages are often low quality and inconsistent with their diffs-known as message-code inconsistency (MCI)<n>We introduce CODEFUSE-COMMITEVAL, the first benchmark designed for MCI detection using large language models (LLMs)<n>We generate seven types of inconsistent messages through rule-guided mutations of originally consistent commits and apply two-fold validation to verify both positive and negative samples.
  arXiv  Detail & Related papers  (2025-11-25T03:33:57Z)
• CLUE: Non-parametric Verification from Experience via Hidden-State Clustering [64.50919789875233]
  We show that correctness of a solution is encoded as a geometrically separable signature within the trajectory of hidden activations.<n>ClUE consistently outperforms LLM-as-a-judge baselines and matches or exceeds modern confidence-based methods in reranking candidates.
  arXiv  Detail & Related papers  (2025-10-02T02:14:33Z)
• MetaFaith: Faithful Natural Language Uncertainty Expression in LLMs [66.14178164421794]
  We introduce MetaFaith, a novel prompt-based calibration approach inspired by human metacognition.<n>We show that MetaFaith robustly improves faithful calibration across diverse models and task domains, enabling up to 61% improvement in faithfulness.
  arXiv  Detail & Related papers  (2025-05-30T17:54:08Z)
• On the Effectiveness of LLMs for Manual Test Verifications [1.920300814128832]
  This study aims to explore the use of Large Language Models (LLMs) to produce verifications for manual tests. Open-source models Mistral-7B and Phi-3-mini-4k demonstrated effectiveness and consistency comparable to closed-source models. There were also concerns about AI hallucinations, where verifications significantly deviated from expectations.
  arXiv  Detail & Related papers  (2024-09-19T02:03:04Z)
• Exploring Automatic Cryptographic API Misuse Detection in the Era of LLMs [60.32717556756674]
  This paper introduces a systematic evaluation framework to assess Large Language Models in detecting cryptographic misuses. Our in-depth analysis of 11,940 LLM-generated reports highlights that the inherent instabilities in LLMs can lead to over half of the reports being false positives. The optimized approach achieves a remarkable detection rate of nearly 90%, surpassing traditional methods and uncovering previously unknown misuses in established benchmarks.
  arXiv  Detail & Related papers  (2024-07-23T15:31:26Z)
• LLMs as Factual Reasoners: Insights from Existing Benchmarks and Beyond [135.8013388183257]
  We propose a new protocol for inconsistency detection benchmark creation and implement it in a 10-domain benchmark called SummEdits. Most LLMs struggle on SummEdits, with performance close to random chance. The best-performing model, GPT-4, is still 8% below estimated human performance.
  arXiv  Detail & Related papers  (2023-05-23T21:50:06Z)
• Contrastive Error Attribution for Finetuned Language Models [35.80256755393739]
  noisy and misannotated data is a core cause of hallucinations and unfaithful outputs in Natural Language Generation (NLG) tasks. We introduce a framework to identify and remove low-quality training instances that lead to undesirable outputs. We show that existing approaches for error tracing, such as gradient-based influence measures, do not perform reliably for detecting faithfulness errors.
  arXiv  Detail & Related papers  (2022-12-21T02:28:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.