Searcharxiv⌕ Search

arXiv subjects

Pingchuan Ma

Publications and source records attributed to Pingchuan Ma.

At least 19 recordsLinked to original sources

QuSema: Detecting Silent Bugs in Quantum Libraries via Quantum-knowledge-enhanced Agents

Quantum libraries are now critical infrastructure for quantum algorithm development, yet their correctness remains difficult to test. Existing testing techniques mainly rely on failure-based or comparison-based oracles, exposing bugs only when executions fail, violate runtime checks, or disagree with another implementation. Their applicability is limited when suitable execution-based oracles are unavailable, leaving some silent bugs undetected. Such missed bugs can produce incorrect results that propagate into experimental conclusions, simulation studies, and algorithmic designs. Here we present QuSema, an autonomous testing agent for finding silent bugs in quantum libraries. QuSema uses constraints from quantum semantics and documentation as a source-level semantic oracle to assess whether implementation logic can produce invalid outputs from valid inputs. It operates through an agentic loop that repeatedly inspects library API documentation and source code, reasons about the intended behavior of quantum operations, identifies potential semantic deviations, and validates them by generating executable tests through library APIs. Guided by quantum-domain reasoning, QuSema turns high-level behavioral mismatches into concrete, user-triggerable bug reports, enabling it to uncover non-crash defects. We implement QuSema for Qiskit and PennyLane. On a benchmark of 20 historical silent bugs, QuSema achieves higher mean bug relocation counts than Claude Code and Codex, with the DeepSeek configuration costing less than Claude Code. QuSema also discovers 40 previously unknown bugs confirmed by the developers, including 30 silent bugs.

cs.SE↗

Inversion-Free High-Dimensional Counterfactual Inference

Counterfactual inference is the task of answering hypothetical "what-if" queries about a unit, given an observational dataset, modeling assumptions, and a causal graph. This enables a modeler to reason about hypothetical scenarios that cannot otherwise be answered without assumptions or a causal graph. The simpler case involves a causal graph with scalar 1-D variables that denote different attributes in the data-generation process. While the literature is clear in the 1-D case, extending this to settings with high-dimensional nodes (e.g., images) remains challenging. Conventionally, in the literature, methods recover the latent state from the factual image (inversion), generate counterfactual attributes from the graph, and render the counterfactual image from the latent state using the updated attributes. In our work, we investigate the necessity of the inversion step in computing high-dimensional counterfactuals. We first discuss moving from the scalar 1-D case to the higher-dimensional case. Secondly, we discuss the relevance of the inversion step and evaluate an inversion-free counterpart repurposed from the image-editing literature. Finally, on image datasets such as CelebA and ADNI, we benchmark our approach using two measures: namely, effectiveness and identity preservation. While effectiveness is a measure of how faithfully the counterfactual attribute is rendered, identity preservation measures whether the same unit (image) is being rendered, both of which are crucial for high-dimensional counterfactual inference. We present the tradeoffs of using inversion-based and inversion-free approaches in our work.

cs.LG↗

TextCloak: Safeguarding Data from Unauthorized LLM Fine-Tuning via RL-Driven Unlearnable Examples

The development of Large Language Models (LLMs) has led to significant advances across a wide range of language tasks, while simultaneously raising growing concerns about unauthorized online data exploitation and privacy leakage. Unlearnable examples (UEs) emerge as a promising defense by injecting crafted perturbations into data such that models trained on them exhibit degraded utility. However, existing methods for text protection are primarily designed for classification tasks and discriminative language models, which limits their broad applicability (e.g., instruction following or reasoning) in LLMs. In this work, we propose **TextCloak**, an RL-driven data synthesis framework that safeguards textual data against unauthorized LLM fine-tuning. TextCloak employs a *UE generative policy* that transforms clean text into unlearnable examples while preserving semantic fidelity and linguistic naturalness. To optimize the policy, we introduce **GRPO-UE**, which simulates the attacker's LLM fine-tuning and rewards downstream degradation via group-relative policy optimization. This overall formulates a bi-level optimization loop, allowing the generator to discover effective protective patterns. Comprehensive experiments on six publicly available datasets and nine state-of-the-art LLMs demonstrate that TextCloak consistently impairs unauthorized LLM fine-tuning while maintaining text utility for legitimate use. Further analyses confirm its transferability and robustness across model architectures, training configurations, and adaptive attacks, highlighting its broad applicability as a practical defense for online content.

cs.CL↗

DPE-Rec: Diversity-Preserving and Efficient Diffusion Language Model Recommendation for Bundle Construction

Generative bundle construction completes a partially observed bundle by generating its missing items from a large product catalog. Compared to autoregressive LLMs, diffusion LLM emerges as a natural fit for this unordered generation task. Their bidirectional attention provides an inductive bias better aligned with unordered bundle construction. Iterative denoising can reveal multiple tokens per step, potentially reducing inference latency. However, directly applying DLMs to this task presents two key challenges: 1 Despite its flexible reveal order, confidence-guided decoding tends to postpone semantic-fork decisions until refinement tokens constrain their choices, reducing bundle diversity. 2 Revealing more tokens per step can compromise recommendation quality, stifling the potential for faster inference. We introduce DPE-Rec, a diffusion framework for diversity preservation and efficient recommendation. DPE-Rec preserves diversity by introducing semantic-fork-aware GRPO, which prioritizes semantic-fork decisions via autoregressive rollouts. It improves efficiency through pseudo-trajectory distillation, which concentrates learning from the vast random corruption space onto the compact trajectory-induced subspace visited by high-quality decoding, unlocking reliable multi-token decoding. Together, these components enable diverse and accurate bundle construction with substantially fewer denoise steps. Experiments on multiple bundle construction benchmarks show that DPE-Rec improves recommendation performance by 63% on average, and its fast variant achieves 25.6$times$ faster inference on average than existing generative recommendation arts, with minimal loss in recommendation quality.

cs.IR↗

Measuring the Permission Gate: A Stress-Test Evaluation of Claude Code's Auto Mode

Claude Code's auto mode is the first deployed permission system for AI coding agents, using a two-stage transcript classifier to gate dangerous tool calls. Anthropic reports a 0.4% false positive rate and 17% false negative rate on production traffic. We present the first independent evaluation of this system on deliberately ambiguous authorization scenarios, i.e., tasks where the user's intent is clear but the target scope, blast radius, or risk level is underspecified. Using AmPermBench, a 128-prompt benchmark spanning four DevOps task families and three controlled ambiguity axes, we evaluate 253 state-changing actions at the individual action level against oracle ground truth. Our findings characterize auto mode's scope-escalation coverage under this stress-test workload. The end-to-end false negative rate is 81.0% (95% CI: 73.8%-87.4%), substantially higher than the 17% reported on production traffic, reflecting a fundamentally different workload rather than a contradiction. Notably, 36.8% of all state-changing actions fall outside the classifier's scope via Tier 2 (in-project file edits), contributing to the elevated end-to-end FNR. Even restricting to the 160 actions the classifier actually evaluates (Tier 3), the FNR remains 70.3%, while the FPR rises to 31.9%. The Tier 2 coverage gap is most pronounced on artifact cleanup (92.9% FNR), where agents naturally fall back to editing state files when the expected CLI is unavailable. These results highlight a coverage boundary worth examining: auto mode assumes dangerous actions transit the shell, but agents routinely achieve equivalent effects through file edits that the classifier does not evaluate.

cs.SE↗

Benchmarking and Enhancing Agentic Causal Discovery

LLM agents can now run causal discovery algorithms on a dataset and report what comes out. Whether they do more than that, such as noticing where an algorithm went wrong and fixing it with further tests, is hard to tell from current benchmarks. Most of them give the variables meaningful names, and an LLM shown only the names of the Tübingen cause-effect pairs already gets 99% of the directions right. With the names removed, the data are usually clean enough that picking a better algorithm than PC closes most of the gap. We introduce Heuresis, a benchmark of anonymous simulated datasets whose mechanisms range from linear Gaussian to heavy-tailed. For each family of datasets it separates what a better existing algorithm can gain from what only a closer look at the data can gain, and it traces every edge an agent changes. Every family leaves a gap that none of the algorithms, the router and the two causal-discovery foundation models we ran can close, so a score above the best chosen algorithm has to come from using the data. We instantiate the benchmark with an agent harness and three LLMs, GPT-6.1 Sol, Claude Opus 5.5 and Qwen 3.8 Flash. GPT-6.1 Sol surpasses the best chosen algorithm on all six families, by 14% adjacency F1 on the most difficult one, and Claude Opus 5.5 does so by 7.4% there. A third of the true edges the two agents restore on the two hardest families appear in no algorithm's output. The code and data will be publicly released to facilitate future work.

cs.LG↗

InstructTA: Instruction-Tuned Targeted Attack for Large Vision-Language Models

Large vision-language models (LVLMs) have demonstrated their incredible capability in visual question answering. However, this rich visual interaction also makes LVLMs vulnerable to adversarial examples. In this paper, we formulate a novel and practical targeted attack scenario that the adversary knows only the vision encoder of the victim LVLM, without the knowledge of its prompts and its underlying large language model. This practical setting poses challenges to the cross-prompt and cross-model transferability of targeted adversarial attack, which aims to confuse the LVLM to output a response that is semantically similar to the attacker's chosen target text. To this end, we propose an instruction-tuned targeted attack (dubbed InstructTA) to deliver the targeted adversarial attack on LVLMs with high transferability. Initially, we utilize a public text-to-image generative model to reverse the target response into a target image, and employ GPT-4 to infer a reasonable instruction $\boldsymbol{p}^\prime$ from the target response. We then form a local surrogate model (sharing the same vision encoder with the victim LVLM) to extract instruction-aware features of an adversarial image example and the target image, and minimize the distance between these two features to optimize the adversarial example. To further improve the transferability with instruction tuning, we augment the instruction $\boldsymbol{p}^\prime$ with instructions paraphrased from GPT-4. Extensive experiments on 6 victim LVLMs demonstrate the superiority of our proposed method in targeted attack performance and transferability. In particular, InstructTA achieves an attack success rate of 51.9% on BLIP-2, outperforming the strongest baseline by 10.5%, and consistently yields the highest attack success rates across all evaluated models. The code is available at https://github.com/xunguangwang/InstructTA.

cs.CV↗

Decompiling Quantum Assembly into Structured Programs

Quantum compilers translate programs into native gates, insert SWAP gates to route interactions onto a device, and optimize the result. The output is quantum assembly, a flat gate list that hides the algorithm's structure (QFT, Grover iteration, QAOA layers). Understanding, auditing, porting, or reusing such assembly requires recovering a program that exposes this structure. We present Quelle, a decompiler that recovers structured programs (library calls, loops, functions, symbolic angles) from routed and optimized quantum assembly (OpenQASM) and checks every output for equivalence with its input. Quelle is built on one invariant: every step is either an exact rewrite, checked where it is applied, or a hypothesis, checked against the input before use. It first lifts the assembly into a circuit by removing compilation artifacts, notably un-routing, which recovers the SWAPs that routing inserted, including those merged into other gates. It then recovers algorithm templates (sketches) whose parameters are solved from quantities preserved by compilation, and, for algorithms without a template, library calls and loops by anti-unification. Failed hypotheses are rejected; the corresponding operations remain in gate-level form. On 816 assembly programs from 22 algorithm families and a random-circuit control, compiled at four routing and optimization levels with native two-qubit gates CX, ECR, or CZ, Quelle's output is equivalent to the input for 100% of the inputs, and it fully decompiles 94% of them; compared on the same inputs, it fully decompiles 1.7x as many as the best LLM-based decompiler, which spends 15K tokens per input. On nine algorithm families for which it has no template, the structure it recovers without templates explains 60% of the gates. It checks inputs of up to 2.3 million gates and emits no inequivalent output on 237 real-world QASMBench assembly files.

cs.PL↗

Well-Typed ComfyUI Cannot Go Wrong: Gradual Refinement Types for Generative-AI Workflow Graphs

Generative-AI workflows are programs assembled from independently developed models and operators. Their correctness depends on relationships among model families, tensor geometry, control modalities, and quantization levels. In ComfyUI, these relationships can be violated even when every connection is accepted by its validator, causing expensive runtime crashes or silent failures. To measure how often such semantic composition failures occur, we classify all 8,766 issues and 1,368 discussion threads of ComfyUI's GitHub repository with an LLM, checked for cross-model agreement. They account for 8\% of issue reports and 21\% of issue-tracker runtime failures, and 32\% of them are silent. Because execution cannot reliably reveal those failures, we present ComfyTy, a gradual refinement type system that checks component compatibility before execution as a compositional property of workflow graphs. Each node declares a reusable schema that states which model families, sizes, modalities, and quantizations it accepts and produces. Wiring nodes together combines their schemas into constraints, which ComfyTy solves to find and report incompatibilities. To evaluate ComfyTy, we first inject faults into official ComfyUI workflows and execute every resulting graph. ComfyTy detects 66.4\% of the confirmed injected faults, all of which ComfyUI's own validator admits. On real workflows across more model families, it raises no false alarm on 765 known-good workflows and detects 67\% of 108 failures rebuilt from user reports. Finally, in a user study with 10 participants, its diagnostics raise fault-localization accuracy on faulty workflows from 35.0\% to 87.5\% and repair accuracy from 30.0\% to 62.5\%.

cs.PL↗

Demistifying Data and Simulator Assumptions in Supervised Causal Discovery

Supervised causal discovery learns to infer causal structure for a new dataset from training datasets paired with structural labels. These training pairs are typically simulated, making the simulator both a source of supervision and a carrier of assumptions about causal graphs, mechanisms, and noise. Understanding the resulting predictions therefore requires examining how these assumptions supplement the information available in observational data, which may be compatible with multiple causal graphs. This paper examines that relationship across representative methods available through June 2026. We organize these methods by prediction target, prediction granularity, encoder, structural decoder, and training regime to relate what each method predicts to how it uses data and simulator-based supervision. Using this framework, we distinguish two questions: whether the target is identifiable under the assumed model class, and whether a trained predictor generalizes beyond its training distribution. Restrictions on mechanisms and noise can make otherwise ambiguous causal directions identifiable, but predictive accuracy under those restrictions does not establish transfer when they change. This distinction motivates evaluation that matches metrics to the identifiable graph target and tests changes in graphs, mechanisms, and noise between training and deployment. Extending such evaluation to real data also requires documenting the external causal evidence and uncertainty behind benchmark reference graphs. Together, these analyses guide method comparison and identify open questions in transfer, test-time adaptation, and uncertainty assessment.

cs.AI↗

efficient matrix-free gauss-newton traveltime tomography on full-metric deformed grids

First-arrival traveltime tomography provides a computationally efficient means of imaging subsurface velocity structure, while boundary-conforming deformed grids allow rugged topography to be represented without abandoning a logically structured mesh. In matrix-free Gauss-Newton inversion, however, the linearized full-metric Eikonal transport and its transpose must be applied repeatedly at a fixed background model. Directional sweeping consequently revisits the same state-dependent dependency structure for every Krylov right-hand side. We recast the frozen tangent transport as a directed graph and distinguish loss of the background-traveltime ordering from genuine algebraic cyclicity. Strongly connected components identify the irreducible part of the transport, whereas the remaining dependencies admit exact scalar substitution after reordering. The condensation graph and local block factorizations are constructed once for each frozen source state and then reused for both primal and transpose applications. Numerical experiments on deformed grids and a three-dimensional tomography problem show that the resulting block-triangular formulation preserves the frozen sensitivity action while substantially reducing the repeated transport cost. By eliminating this dominant inner-iteration cost, the proposed method markedly accelerates Gauss-Newton computation and makes full-metric matrix-free inversion on deformed grids practical at much lower cost.

math.NA↗

Sound Debloating of Redundant Checks in Zero-Knowledge Machine-Learning Circuits

Zero-knowledge (ZK) proof systems for neural-network inference compile the model into a system of arithmetic constraints. Many of these constraints are redundant checks: range proofs, sign lookups, and bit decompositions who are globally entailed by the rest of the circuit through chains of reasoning that span distant gadgets. Removing them shrinks the circuit and accelerates proving, but the removal must be carefully justified: an unsoundly debloated circuit becomes forgeable, accepting witnesses the original would have rejected and so allowing a prover to claim, for example, that a neural network produced an output it never actually computed. Such soundness vulnerabilities are not hypothetical: under-constrained circuits in deployed ZK systems have enabled attackers to forge transactions and bypass verification entirely. We present an automated framework that removes redundant checks while provably preserving soundness. For each candidate removal, our tool first checks whether the rest of the circuit, on its own, can still rule out every value the removed check was excluding. Using whole-circuit abstract interpretation, the analysis searches for such alternative justifications and records them in a provenance graph; a check is then removed only when an alternative path through the graph still derives the facts that it is checking. This ensures that the debloated circuit opens no new forging strategy to an adversary. We evaluate circuits spanning MLP, CNN, RNN, and transformer architectures generated by two production frameworks (ezkl and zkml), with up to 25.3 million constraints. Our tool removes up to 48.7\% of constraints and reduces prover time by up to 72.8\%, without weakening security.

cs.CR↗

ENCORE: Event-Assisted Complementary Motion Refinement for Learned Video Compression

Learned video compression relies on accurate temporal modeling to remove redundancy between adjacent frames. However, most existing codecs infer motion solely from discretely sampled RGB frames, making their estimates vulnerable to fast motion, blur, occlusion, weak texture, low illumination, and abrupt brightness changes. Event cameras asynchronously capture fine-grained intensity changes between RGB timestamps and therefore provide complementary evidence about inter-frame dynamics. We propose ENCORE, an Event-Assisted Complementary Motion Refinement framework for learned video compression. ENCORE first employs Complementary Motion Representation (CMR) to decompose aligned RGB-event features into common and modality-specific motion representations. Spatial Energy and Redundancy-Informed Calibration (SERIC) then identifies event-specific responses that are active and novel relative to RGB, suppresses weak or redundant evidence, and predicts a candidate flow correction. Finally, Energy-Aware Routing (EAR) determines where and how strongly the correction should refine the RGB flow. Events serve solely as an auxiliary modality for motion modeling, while RGB remains the only coding and reconstruction target. Experiments on BS-ERGB, HQ-EVFI, and CED demonstrate consistent gains across datasets and GOP lengths. On BS-ERGB, ENCORE achieves up to 20.80% PSNR-RGB and 22.14% MS-SSIM-RGB BD-rate savings, while retaining clear improvements on the other two datasets.

cs.CV↗

KQFuzz: Knowledge-Guided Fuzzing for Quantum Libraries via Large Language Models

As quantum computing continually improves, ensuring the reliability and correctness of quantum libraries has become increasingly critical. To this end, many LLM-based fuzzing approaches towards quantum libraries have been proposed to uncover potential bugs. However, these methods still suffer from limitations such as insufficient flexibility and low efficiency, which hinder the progress of the quantum computing field. To address these challenges, we propose KQFuzz, a novel knowledge-guided fuzzer for quantum libraries. It leverages comprehensive codebase knowledge to ground LLM-based test generation, synergizing this with fitness-guided evaluation and two-level mutations to explore complex execution paths and trigger potential bugs. Firstly, KQFuzz introduces a novel prompting scheme tailored to quantum programs, which strategically incorporates knowledge of the codebase to efficiently generate high-quality quantum seed programs. Moreover, we develop evaluation and mutation strategies to handle the generated seed programs, facilitating efficient fuzzing execution while further enriching the diversity of the resulting test cases. We implement KQFuzz and conduct fuzzing on three popular quantum libraries, including Qiskit, PennyLane, and Cirq. Experimental results demonstrate that our approach significantly outperforms other state-of-the-art methods, with coverage improved by up to 18.44%. During the development of KQFuzz, we discovered 13 bugs, all of which have been confirmed and 12 have already been fixed by the developers.

cs.SE↗

OmniQEC: discovering practical quantum error-correcting codes by an AI scientist

Quantum error correction (QEC) is indispensable for scalable fault-tolerant quantum computing. However, discovering QEC codes that remain effective is challenging, as logical performance depends on the interplay between code structure, hardware, syndrome extraction, and decoding, which often impose competing requirements. Here we introduce OmniQEC, an efficient AI scientist for discovering QEC codes suited to deployment on modern quantum processors. OmniQEC formulates QEC design as an iterative discovery process in which an orchestrator, implemented by advanced large language models (LLMs), coordinates code generation, code-level screening, syndrome-extraction synthesis, and decoder-based circuit evaluation. At its core, OmniQEC combines a self-evolving reasoning mechanism with a slow--fast synergistic workflow: a fast loop explores candidates using inexpensive code-level proxies, whereas a slow loop performs physically grounded circuit-level evaluation and feeds the resulting evidence back into the search. We evaluate OmniQEC across four qLDPC construction families, three LLM backends, and $14$ total-physical-qubit budgets per backend. The discovered codes show steadily improving logical-error suppression with increasing physical-qubit budgets and outperform the BB codes with $[\![72,12,6]\!]$ and $[\![144,12,12]\!]$ under complete-implementation budgets of 98 and 240 physical qubits, respectively. The discovered codes are hardware-friendly and may be of independent interest for practical QEC implementation. These findings pave the way towards LLM-assisted QEC discovery grounded in physically informed code--circuit--decoder co-design.

quant-ph↗

Contrastive-Augmented Flow Matching for Style-Content Disentanglement

Learning representations that separate content and style is crucial for controllable generation and compositional generalization. However, diffusion and flow-based models trained primarily with generative objectives often produce entangled or misaligned factors. To address this gap, we introduce Contrastive Augmented Flow Matching (CAtFM), a framework that integrates contrastive regularization into an invertible flow matching formulation to promote structured content-style representations. Rather than constraining intermediate latents or velocity fields, we apply contrastive supervision to predicted endpoints during training, enforcing semantic consistency across transported distributions while allowing disentanglement to emerge implicitly, without assuming strictly pure or fully factorized content and style representations. Our main experiments operate in the CLIP embedding space, with additional validation using frozen DINO and ALIGN encoders. Across synthetic data, in-domain styles, and real-world benchmarks (ImageNet, WikiArt, DomainNet, and DTD), CAtFM improves content and style retrieval, enhances embedding cluster separation, and achieves stronger open-set robustness compared to generative and discriminative baselines. Overall, CAtFM provides a simple way to couple discriminative constraints with deterministic transport, improving disentanglement and robustness under distribution shift.

cs.CV↗

AutoSpec: Safety Rule Evolution for LLM Agents via Inductive Logic Programming

Large language model (LLM) agents increasingly automate complex tasks by integrating language models with external tools and environments. However, their autonomy poses significant safety risks: agents may execute destructive commands, leak sensitive data, or violate domain constraints. Existing safety approaches face a fundamental tradeoff: hand-crafted rules are interpretable but brittle, with overly conservative rules blocking safe operations (high false positives) while permissive rules miss unsafe behaviors (high false negatives). Neural classifiers lack the interpretability required for safety-critical deployments. We present AutoSpec, a framework that automatically evolves deployed expert-designed safety rules from user safe/unsafe annotations through counterexample-guided inductive synthesis (CEGIS) guided by inductive logic programming (ILP). Starting from the expert rules and a stream of annotated traces, AutoSpec iteratively evaluates rules, mines false-positive and false-negative counterexamples, uses ILP to learn which predicates discriminate them, generates candidate rule edits, and verifies candidates to select the best revision. The key insight is that ILP efficiently identifies predicates that appear frequently in false negatives but rarely in false positives (or vice versa), dramatically pruning the exponential search space of rule edits. This continues until convergence, producing interpretable rules that balance precision and recall. We evaluate AutoSpec on 291 execution traces spanning code execution and embodied agent domains. AutoSpec raises rule F1 to 0.98 and 0.93 across the two domains, achieving up to 94% false positive reduction while maintaining high recall, and converges within 4-5 iterations. The ILP-guided approach achieves up to 4.8x higher F1 than heuristic CEGIS. The learned rules are human-readable, auditable, and generalize to unseen scenarios.

cs.SE↗

SkillReducer: Optimizing LLM Agent Skills for Token Efficiency

LLM-based coding agents rely on \emph{skills}, pre-packaged instruction sets that extend agent capabilities, yet every token of skill content injected into the context window incurs both monetary cost and attention dilution. To understand the severity of this problem, we conduct a large-scale empirical study of 55,315 publicly available skills and find systemic inefficiencies: 26.4\% lack routing descriptions entirely, over 60\% of body content is non-actionable, and reference files can inject tens of thousands of tokens per invocation. Motivated by these findings, we present \textsc{SkillReducer}, a two-stage optimization framework. Stage~1 optimizes the routing layer by compressing verbose descriptions and generating missing ones via adversarial delta debugging. Stage~2 restructures skill bodies through taxonomy-driven classification and progressive disclosure, separating actionable core rules from supplementary content loaded on demand, validated by faithfulness checks and a self-correcting feedback loop. Evaluated on 600 skills and the SkillsBench benchmark, \textsc{SkillReducer} achieves 48\% description compression and 39\% body compression while improving functional quality by 2.8\%, revealing a \emph{less-is-more} effect where removing non-essential content reduces distraction in the context window. These benefits transfer across five models from four families with a mean retention of 0.965, and generalize to an independent agent framework.

cs.SE↗