{"total":15,"items":[{"citing_arxiv_id":"2607.05936","ref_index":11,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Mitigating Errors in LLM-Generated Web API Invocations via Retrieval-Augmented Generation and Constrained Decoding","primary_cat":"cs.SE","submitted_at":"2026-07-07T07:38:28+00:00","verdict":"CONDITIONAL","verdict_confidence":"HIGH","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Constrained decoding derived from OpenAPI specifications eliminates hallucinated web API calls in LLM-generated code and substantially improves correctness across 24 models, while retrieval-augmented generation yields inconsistent, model-dependent gains.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2607.05810","ref_index":33,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"SCOPE: Leveraging Subgoal Critiques for Code Generation","primary_cat":"cs.SE","submitted_at":"2026-07-07T04:09:41+00:00","verdict":"CONDITIONAL","verdict_confidence":"MODERATE","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A Lean-oriented prover model, fine-tuned with dense and sparse RL rewards, generates structured semantic critiques that improve LLM code generation accuracy over Reflexion and Self-Refine on LiveCodeBench V6 and BigCodeBench.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09800","ref_index":9,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"FASE: Fast Adaptive Semantic Entropy for Code Quality","primary_cat":"cs.SE","submitted_at":"2026-06-08T17:53:05+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"FASE approximates functional correctness via MST on structural and semantic dissimilarity graphs, reporting 25% better Spearman correlation and 19% better ROCAUC than LLM-based semantic entropy at 0.3% runtime cost on HumanEval and BigCodeBench.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.08135","ref_index":9,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TICoder: A Repository-Level Code Generation Framework with Test-Driven Planning and Implementation-Aware Reuse","primary_cat":"cs.SE","submitted_at":"2026-06-06T12:23:30+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"TICoder improves repository-level code generation by 11.52% over prior methods through test-driven planning and implementation-aware code reuse on standard benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29372","ref_index":18,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"On the Road to Personalized Code Intelligence: Portraiting and Assisting Developers Based on Their In-IDE Behaviors","primary_cat":"cs.SE","submitted_at":"2026-05-28T05:17:08+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"VirtualME is a new infrastructure that continuously extracts and interprets in-IDE developer behaviors to build personalized personas, delivering 33.8% better performance on repository-level knowledge Q&A than generic baselines.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.17029","ref_index":11,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Task Abstention for Large Language Models in Code Generation","primary_cat":"cs.SE","submitted_at":"2026-05-16T14:58:11+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A distribution-free abstention rule grounded in multiple hypothesis testing uses execution consistency to let code LLMs avoid hallucination-prone tasks with theoretical guarantees.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.15503","ref_index":8,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"uGen: An Agentic Framework for Generating Microarchitectural Attack PoCs","primary_cat":"cs.CR","submitted_at":"2026-05-15T00:50:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"uGen is the first retrieval-augmented multi-agent LLM framework for generating functionally correct microarchitectural attack PoCs, reporting up to 100% success on Spectre-v1 and 80% on Prime+Probe at low cost.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.26672","ref_index":55,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"What Makes Software Bugs Escape Testing? Evidence from a Large-Scale Empirical Study","primary_cat":"cs.SE","submitted_at":"2026-04-29T13:42:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Post-release defects concentrate in older, frequently modified high-churn components and require longer and more complex fixes than pre-release defects.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.10126","ref_index":19,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"MR-Coupler: Automated Metamorphic Test Generation via Functional Coupling Analysis","primary_cat":"cs.SE","submitted_at":"2026-04-11T09:42:44+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"MR-Coupler leverages functional coupling analysis and LLMs to generate valid metamorphic test cases for over 90% of tasks while detecting 44% of real bugs, outperforming baselines by 64.90% in validity and 36.56% in false-alarm reduction.","context_count":1,"top_context_role":"background","top_context_polarity":"support","context_text":"This structured prompt provides both contextual information (ii-v) and task description (i and vi), guiding the LLM to generate syntactically correct MTCs. The details of employed LLMs and their configuration can be found in Section 4.2.1. The output of this step is a set of candidate MTCs, each implemented as a standalone test class. 3.2.2 MTC Refinement.Consistent with prior observations [ 19, 46, 66], LLM-generated code frequently fails to execute, commonly due to errors such as \"cannot find symbol\". These errors typically arise from two sources: (i) referencing non-existent classes, APIs, or fields, or (ii) missing dependencies (e.g., absent import statements). To deal with this,MR-Couplerrefines each non- compilable or non-executable MTC."},{"citing_arxiv_id":"2603.06276","ref_index":25,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Story Point Estimation Using Large Language Models","primary_cat":"cs.SE","submitted_at":"2026-03-06T13:34:09+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LLMs predict story points better in zero-shot prompting than supervised deep learning models trained on 80% of project data, with few-shot examples and comparative judgments further improving performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.22202","ref_index":13,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Library Hallucinations in LLM-Generated Code: A Risk Analysis Grounded in Developer Queries","primary_cat":"cs.SE","submitted_at":"2025-09-26T11:14:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"A study of seven LLMs finds that realistic prompt variations such as one-character misspellings trigger library hallucinations in up to 26% of cases, fabricated names in up to 99%, and time-based prompts in up to 85%, and introduces LibHalluBench for evaluation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2505.04080","ref_index":128,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"MojoFrame: Dataframe Library in Mojo Language","primary_cat":"cs.DB","submitted_at":"2025-05-07T02:42:05+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MojoFrame is the first dataframe library in Mojo, using tensors for numeric columns and cardinality-aware methods for others, with up to 4.6x speedup on TPC-H queries versus other libraries.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"a relational schema and use tensor operations only after the query planning stage. f) Systems for Speeding up Data Science Coding:There exist works for speeding up the coding process for building data science pipelines [122]-[130]. Code completion tools recommend next lines of code for the user via either rule- based [125], [126] or LLM-based [127], [128] predictions. Checkpointing tools such as Diff-in-the-loop [122], Elastic- Notebook [123], [131], Kishu [124], [132], [133], and Chip- mink [134] can be used to save states of data science pipelines for returning to later, facilitating more efficient code iteration. Symphony [129] and B2 [130] adopt a non-coding approach and enable point-and-click interactions with ML models and"},{"citing_arxiv_id":"2408.15815","ref_index":9,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"MR-Adopt: Automatic Deduction of Input Transformation Function for Metamorphic Testing","primary_cat":"cs.SE","submitted_at":"2024-08-28T14:24:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MR-Adopt deduces input transformations from hard-coded MR test cases using LLMs, data-flow refinement, and output-relation selection to enable reuse with new source inputs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2403.17134","ref_index":35,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"RepairAgent: An Autonomous, LLM-Based Agent for Program Repair","primary_cat":"cs.SE","submitted_at":"2024-03-25T19:17:43+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"RepairAgent autonomously repairs 164 bugs on Defects4J including 39 not fixed by prior techniques by treating an LLM as an agent that invokes tools via a finite state machine and dynamic prompts.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2402.19473","ref_index":241,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Retrieval-Augmented Generation for AI-Generated Content: A Survey","primary_cat":"cs.CV","submitted_at":"2024-02-29T18:59:01+00:00","verdict":"ACCEPT","verdict_confidence":"MODERATE","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A survey classifying RAG foundations for AIGC, summarizing enhancements, cross-modal applications, benchmarks, limitations, and future directions.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"version of the \"next sentence prediction\" task. Query-based RAG is the mainstream paradigm for code completion. Drain et al. [240] retrieved template functions for function completion. ReACC [91] uses both sparse and dense retrieval. RepoCoder [186] performs iterative RAG by augmenting the retrieval input with previously generated code. De-Hallucinator [241] retrieves API references using first-time generated contents, then conducts query-based RAG for improved code completion. REPOFUSE [242] includes rationale context and retrieved codes to form prompt, and ranks the contexts to fit in the length limit. Many works leverage latent representation-based RAG. Retrieve-and-edit [100], RepoFusion [243], and EDITAS [244]"}],"limit":50,"offset":0}