Jul 2026· International Conference on Automated Software Engineering· Vol 33· 0 citations· 47 references
Computer Science
TL;DR
Results show that AGR achieves strong decision accuracy in triage and evidence verification, and produces more actionable and engineering-useful issue specifications than both raw feedback and a strong LLM baseline, while reducing unsupported details.
Abstract
Post-release user feedback is a major control signal for maintenance and evolution in modern software development, yet it is noisy, fragmented, and difficult to translate into developer-usable issue specifications. Large Language Models (LLMs) can assist this transformation, but they often hallucinate or over-commit when evidence is weak, conflicting, or incomplete, limiting their robustness in automated software engineering workflows. We propose AGR (Agent-Governed Retrieval-Augmented Generation), a framework that regulates evidence acquisition and generation decisions via agentic control. AGR first applies an agentic triage step to filter low-signal or off-topic feedback, then retrieves evidence from a three-category hierarchy comprising official documentation, historical bug reports, and targeted web sources. It further performs confidence-weighted fusion across authoritative categories and uses an agentic decision module to verify relevance and sufficiency, trigger additional retrieval or online search when needed, reuse prior reports via memory, and abstain when evidence-supported grounding cannot be established. We evaluate AGR on two open-source software ecosystems, Firefox and VS Code. Results show that AGR achieves strong decision accuracy in triage and evidence verification, and produces more actionable and engineering-useful issue specifications than both raw feedback and a strong LLM baseline, while reducing unsupported details.
Repository-level code generation relies on heterogeneous evidence whose relevance, compatibility, and completeness are inherently uncertain. Similar-code examples, repository context, and project-specific APIs may provide complementary information, but can also introduce noisy, redundant, or conflicting signals. Existing retrieval-augmented approaches primarily optimize retrieval relevance without explicitly modeling how uncertainty in retrieved evidence affects downstream generation. We introduce OpenCoder, an uncertainty-aware framework that estimates source-specific uncertainty, uses it to filter and rank heterogeneous evidence, and guides generation, verification, and repair. A factorial analysis over API knowledge, repository context, and similar-code evidence reveals no universal additive source ranking; instead, significant cross-source interactions depend on the accompanying evidence and LLM backend. On an expanded 32-task RepoExec-inline evaluation, OpenCoder improves GPT selected-output correctness over Baseline RAG from 56.25\% to 78.13\%. However, it matches a verification-and-repair control, and the corresponding Gemini improvement is not statistically supported, indicating backend-dependent benefits. Target-aware API refinement also substantially improves API-set retrieval. These findings support treating uncertainty as an actionable control signal for repository-level retrieval, verification, and repair.
Chandan Kumar Sah, Xiaoli Lian, Li Zhang· arXiv.org· 1 citation
A controlled seed-intervention pilot finds that retrieval-derived initial context yields higher file F1 with less post-seed exploration than random non-gold context, while oracle gold context shows substantial remaining headroom.
The paper contributes three reusable deployment patterns: hybrid RAG evidence construction, multi-channel retrieval and reranking produce auditable FAQ candidates, and trace-driven RAG and reranker improvement, where reranker fine-tuning is evaluated not only for in-domain gain but also for forgetting risk.
Existing long-video agents acquire evidence through one uniform behavior, ignoring whether the required evidence is concentrated, requires broad occurrence coverage, or must discriminate competing hypotheses---which can cause failure before substantive reasoning begins. Prescribing a fine-grained solution procedure for every question is not a satisfactory remedy, as it restricts autonomous exploration. We propose VESTA, a training-free long-video agent organized as a route-conditioned acquire--verify--consolidate loop. Before exploration, an intent router infers an evidence-acquisition policy---focused, recall, or contrastive retrieval over a shared visual--speech scene index---together with an evidence-accounting policy that configures the evidence view maintained during exploration. Policy-steered retrieval yields provisional references that multimodal evidence operations convert into observations, while the Reasoner remains free to verify them, re-query using intermediate findings, or inspect regions outside the retrieved set. A temporal evidence ledger consolidates observations into an adaptive, compressed view of temporal location, provenance, coverage, conflicts, verification outcomes, and hypothesis support, exposing missing and unresolved evidence to guide subsequent acquisition; finalization prioritizes verified observations. On Video-MME-v2, VESTA improves average accuracy by 2.7 points over VideoARM and gains across all six reported metrics. On LongVideoBench, EgoSchema, and LVBench under shared query-time models, it improves by 6.9 points on the LongVideoBench long subset and 1.5 on LVBench, and matches VideoARM on EgoSchema.
Can-Can Zhang, Bao-Feng Zhang, Xiao-Tian Han et al.· 0 citations
Product catalogs are the backbone of e-commerce sites, yet a large number of structured attributes (SAs) -- such as material, color, and shape -- often have missing values. Typically, SA values are extracted from product information, including titles and descriptions. While LLM-based generator-evaluator frameworks have demonstrated effectiveness for SA prediction -- where an LLM generates SA values and another evaluates them -- they face challenges when the Generator and Evaluator produce conflicting outputs, as either component can make mistakes. We introduce \texttt{CatalogAgent}, a novel agentic system that continuously improves Generator and Evaluator models for e-commerce catalog enrichment. When disagreements arise from (1) internal conflicts between the LLM-based Generator and Evaluator, or (2) external feedback from sellers on LLM outputs, a Supervisor Agent intervenes to mediate these conflicts and make final decisions. The system also incorporates a Memory Base and a Memory Summarizer that stores Supervisor Agent activities from individual cases and aggregates patterns into learnings. These learnings are fed back to the worker Generator and Evaluator LLMs, enabling self-improvement without human intervention. Through context engineering -- injecting learnings and insights into worker LLMs'contexts -- the system successfully transfers the Supervisor's capabilities to the Generator and Evaluator, improving their performance by 15.24\% and 13.98\%, respectively. Our experiments demonstrate a new paradigm of Supervisor Agent-mediated self-learning systems for improving generative AI model accuracy.
The Assistant-Scribe-Knowledge Checker (ASK) framework for generating structured specification documents through guided interviews is evaluated in a consulting-firm setting where consultants are required to produce project “return-of-experience” documents to capture reusable knowledge.
Sylvain Roudiere, Bianca Lento· European Conference on Knowl...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
