公开论文雷达

Agent harnesses record a failed tool call and its error message in the transcript and ask the model to continue, on the assumption that the error is corrective information. We measure whether it is. Defining the corrective gain of a failure record as the change in log-probability of re-emitting the action that just failed, we find the gain is negative for every instruction-tuned model we tested (6 checkpoints, 135M-1.7B, 4 families) in two environments: simulated tool calling and MBPP program repair. Normalised by action length the effect is about -1.03 nats per action token, a factor of 2.8 in the odds of each token, and holds on 90%-100% of individual items, not only on average. Over a fixed candidate set the probability of repeating the failed call rises from 0.06 to 0.54, and greedy decoding reproduces it token for token on 19% of items after the failure versus 0% before. Counterfactuals pairing the same call with a failure message, a success message, or a neutral acknowledgement separate two effects: the failed call's surface form accounts for 83% of the damage, while the semantic contribution of marking it failed is small and inconsistent in sign across environments. The problem is in the harness, not the model's grasp of error messages, and that predicts which remedies work. Replacing the verbatim call with a runtime-generated description of the failure removes 76% of the inversion at no token cost, and making previously-failed strings unreachable at the decoder acts on the same term. Two plausible remedies do not: an explicit "do not repeat" instruction leaves the measured quantity where it was, and deleting the failed attempt to retry from a clean context, the standard prescription for context contamination, is the worst harness we measured for repetition, because it restores the context that produced the failure. The study runs end to end on a CPU; all artefacts are released.

Large Language Models (LLMs) offer new opportunities for automated code refactoring. However, generated changes must reduce targeted quality problems without introducing new issues or altering behaviour-relevant code structures. We introduce REFINE (Refactoring with Evidence-aware Flow for Integrated ageNtic Execution), a tool-agnostic, evidence-aware multi-agent approach for generating Java file-level refactoring candidates. REFINE combines static-analysis-guided smell identification, smell-informed planning, LLM-based transformation, automated re-analysis, preservation checks, and structured reporting. We evaluate REFINE on 450 Java files from 15 open-source systems, producing 1,350 model-pass outputs using OpenAI GPT-5.5, Google Gemini 3.1 Pro Preview, and Anthropic Claude Opus 4.8. REFINE reduces detected code smells by 68.26%, 72.79%, and 68.49% across the three configurations, respectively, with the strongest reductions observed for major smells. A matched 150-file direct-prompt baseline shows that REFINE achieves a higher median code-smell reduction with smaller edits and fewer public-method removals. However, broader quality improvements are inconsistent, and preservation checks reveal residual risks, including assert/fail-call changes and public-method removal. Therefore, REFINE outputs should be treated as refactoring candidates requiring compilation, testing, dependency analysis, and human review before adoption in repository- or system-level settings.

Requirements elicitation is essential for developing interactive software systems, as it helps ensure that the resulting product meets stakeholder needs. Since elicitation typically relies on natural language (NL), misunderstandings can arise from its inherent ambiguity. Formal specifications can reduce ambiguity but require technical expertise. GUI prototyping therefore provides a valuable alternative by turning requirements into tangible visual artifacts that support communication, elicitation, and validation. However, creating high-fidelity prototypes remains time-consuming and costly. Similarly, requirements verification, which ensures that implementations conform to specified requirements, is still largely manual, while existing automated approaches are often limited to static, rule-based techniques. This work addresses two challenges: (C1) reducing the effort required to transform NL requirements into GUI prototypes, and (C2) reducing the effort required for requirements verification in GUI applications and prototypes. For C1, we introduce novel NL-based GUI retrieval and reranking methods, new benchmarks, and techniques for efficiently adapting LLMs to GUI generation, including proprietary GUI representations. Their effectiveness is demonstrated on a large benchmark with human annotations. For C2, we propose LLM-based methods for verifying semantically complex NL requirements on static GUI prototypes and introduce a multimodal LLM-based agent for verifying complex functional and non-functional requirements in dynamic GUI applications through automatically generated and evaluated interaction trajectories. Overall, the proposed methods substantially reduce manual effort in GUI prototyping and requirements verification.

3D geometry editing is a critical yet labor-intensive part of the graphics pipeline, requiring artists to translate creative intent into precise operations in complex professional software. Large language models (LLMs) have shown promise for script-based 3D creation, but script generation is less suited to perception-driven editing of arbitrary existing meshes, where execution must remain visually grounded and untouched regions should be preserved. We present a \emph{visual-centric}, training-free multi-agent system that edits existing 3D meshes directly in Blender by emulating the iterative workflow of human artists. Rather than generating scripts or regenerating geometry, our system operates through the Blender GUI: multimodal LLM agents observe the viewport, reason about the current mesh state, and execute localized edits through simulated user interactions. Experiments on a curated benchmark provide initial evidence that this agentic approach can follow natural language instructions, perform representative localized mesh edits, and preserve the overall identity of the input asset. Our results highlight a complementary regime for language-driven 3D editing: direct in-place modification of existing meshes within the native 3D editing workflow. We view this work as an exploratory step toward visual-centric agentic geometry editing in professional graphics software.

Visual foundation models are commonly adapted under the assumption that the appearance of incoming data may change while the semantic meaning of the prediction task remains fixed. In long-lived visual systems, however, taxonomies, policies, and concept definitions can themselves evolve, causing the same visual evidence to require a different interpretation. We study this setting as evolving semantic concept shift and introduce SemReWrite, a framework for selectively updating obsolete visual--semantic mappings while preserving knowledge that remains valid. SemReWrite represents changes between old and revised semantic specifications, combines semantic discrepancy with sparse revised supervision to localize affected visual regions, and uses an input-dependent low-rank rewriting mechanism together with structured semantic memory, preservation, and obsolete-decision suppression. We further introduce EvoShift-Bench, spanning ImageNet, iNaturalist, CUB-200-2011, and DomainNet, with semantic transitions including class split, merge, boundary revision, insertion, partial redefinition, recurrence, and mixed semantic--appearance shift. To explicitly evaluate selective semantic revision, we introduce Rewrite Accuracy (RA) and Preservation Accuracy (PA) for affected and unaffected regions, respectively, Obsolete Retention (OR) for measuring residual outdated semantic associations, and the Selective Revision Score (SRS), which jointly summarizes rewriting and preservation performance. Experiments show that SemReWrite achieves a stronger balance between learning revised semantics and retaining unaffected knowledge than prompt replacement, conventional fine-tuning, parameter-efficient adaptation, and continual-learning strategies.