PyTorchKR Papers

최신 논문

총 568,830편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2610.14381v2cs.AI

ACED-Bench: Evaluating How LLM Agents Acquire and Act on Causal Evidence

Large language models are increasingly used as scientific assistants, yet we still lack reliable ways to evaluate whether they can investigate a system rather than merely answer a prompt. A useful evaluation must let agents choose experiments, interpret finite evidence, and make causal decisions, while preserving objective ground truth. We introduce ACED (Active Causal Evidence and Decision Benchmark), a neuro-symbolic benchmark that separates the readable scientific surface from the formal causal world: natural-language scenarios describe the system, while a hidden causal model defines the data, the effect of every intervention, and the oracle answers. In ACED-Repair, agents diagnose a failing production system: within 15 experiments they must decide which controls to test, read confounded and filtered data, and recommend a fix, which we re-simulate on a hidden true objective. In ACED-Query, agents answer causal questions and make constrained causal decisions in worlds backed by categorical probabilistic graphical models. Acceptance on ACED-Repair ranges from 43% to 0% across 12 LLMs, while strong agents approach ceiling on ACED-Query's qualification questions. Tracing every experiment shows where investigations stop: most open-weight models test a true lever on its own in fewer than half of the worlds, the strongest models test the right levers but often miss the working dose or combination, and agents that do run a working fix almost always recommend it. No model reliably repairs systems with two independent causes or systems that need a subgroup-specific policy. These results suggest that reliable AI co-scientists need explicit support for choosing experiments, keeping track of evidence, and composing interventions. Code and data are available at https://github.com/Rose-STL-Lab/ACED-Bench.

arXiv:2610.13740v2cs.RO

SmolWAM: 엣지 디바이스 로봇 조작을 위한 잠재 월드-액션 모델

SmolWAM: A Latent World-Action Model for Robotic Manipulation on Edge Devices

World-action models (WAMs) jointly predict robot actions and future observations, but their video diffusion backbones are costly for edge deployment. We introduce SmolWAM, a compact latent WAM derived from the 2B-parameter Cosmos Policy using single-frame encoding, cached text features with preserved attention sinks, depth pruning, and token compaction. With 12 of 28 transformer blocks and 1,008 tokens, SmolWAM predicts actions and future latents in one rectified-flow step. After 3.6 A100 GPU-hours of training, it achieves 95.5% success over 400 LIBERO episodes at 57ms, 2.3TFLOP, and 2.4GiB per 16-action chunk on an A100: 11x lower latency and 3.2x lower memory than the five-step teacher. Its world head predicts scene motion; joint training improves success by 6.2 percentage points over an action-only ablation.

arXiv:2610.13277v3cs.IR

Beyond the Final Embedding: Retrieving Memories from Token-Level Evidence

Memory-augmented assistants retrieve evidence from stored conversations and documents to answer queries. Most embedding-based retrievers compare single-vector representations of queries and memories, even when relevance depends on a small part of the content. Such local evidence can be matched to queries using the contextual token representations already computed by the same embedding models. Our method uses sparse token matching to improve retrieval without additional training. Each query token retrieves a small set of matching tokens across candidate memories. The matching scores are aggregated for each memory and combined with whole-memory similarity to rank memories. An optional low-rank query adapter can be trained with supervision while keeping the encoder and stored memory representations fixed. Our method improves retrieval across eight benchmarks and multiple model families without additional training. Query adaptation further increases the gains to 13.3 and 14.3 percentage points in average Recall@10 and NDCG@10 on Qwen3-Embedding-4B. These retrieval gains translate into higher question-answering accuracy on long-term memory benchmarks.

arXiv:2610.12468v2cs.RO

DreamTrue: 반사실적 사후 학습을 통한 행동 충실형 로봇 월드 모델

DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training

본 논문에서는 행동 충실성(action-faithful)과 물리적 타당성을 갖춘 비디오 예측을 위한 다중 뷰, 크로스 임바디먼트(cross-embodiment) 로봇 월드 모델인 DreamTrue를 제시합니다. 기존 로봇 데이터셋에서 이러한 모델을 학습하는 데는 두 가지 장애물이 존재합니다. 부정확한 캘리브레이션은 행동 추종(action following) 능력을 저하시킬 수 있으며, 실패한 상호작용에 대한 제한적인 커버리지는 예측이 성공적인 결과로 치우치는 바이어스를 유발할 수 있습니다. 다양한 임바디먼트 전반에 걸친 행동 추종 성능을 향상시키기 위해, 행동 궤적을 이미지 공간 조건으로 렌더링하고 이러한 조건을 타깃 비디오와 정렬하기 위한 오프라인 기하학적 캘리브레이션을 도입합니다. 또한 상호작용 커버리지를 넓히기 위해 기록된 행동 궤적을 수정하고 더 넓은 범위의 행동 및 접촉 구성 하에서 미래 비디오를 생성하는 반사실적 사후 학습을 도입합니다. 쌍을 이루는 실제 정답(ground-truth) 미래 비디오 없이도 이러한 예측에 대한 피드백을 제공하기 위해, 로봇, 객체, 상호작용 결함을 포괄하는 인간 어노테이션 비디오 데이터셋을 구축하고 이를 활용해 임바디드 비디오 보상 모델을 학습합니다. 이 보상 모델의 점수는 강화학습 사후 학습을 물리적으로 더욱 타당한 상호작용 결과로 유도합니다. AgiBot에서 DreamTrue는 SOTA(state-of-the-art) 행동 추종 성능을 달성하는 동시에, 인간 평가 상호작용 결함률을 48.12%에서 6.25%로 줄였습니다. 주목할 점은 본 모델이 AgiBot World Challenge 2026의 월드 모델 트랙에서 1위를 차지했다는 것입니다. 프로젝트 페이지는 https://brave-eai.github.io/DreamTrue 에서 확인할 수 있습니다.

arXiv:2610.11288v2cs.LG

랜덤 특징(Random Feature)을 사용한 조건부 디퓨전 모델에서의 암기와 악성 일반화

Memorization and Malign Generalization in Conditional Diffusion Models with Random Features

Conditional diffusion models generate diverse, novel, and high-quality samples under prescribed conditions. However, theoretical understanding of their memorization and generalization remains limited, while recent works have characterized these behaviors primarily in unconditional settings. In this work, we analyze a random-feature conditional score model in the high-dimensional proportional limit, deriving asymptotic expressions for training and test losses. By decomposing the test loss, we show that in the overparameterized regime, increasing model width improves prediction of the condition-dependent mean while reducing within-condition prediction variance, a phenomenon we term "malign generalization." We further show that in the infinite-width limit, the model predicts the conditional mean as accurately as ridge regression of the data on the condition, while its within-condition variance vanishes. Finally, analyzing the training loss reveals that more informative conditions lead to memorization of training samples at smaller widths. These theoretical findings are supported by experiments with U-Net architectures on realistic data.

arXiv:2610.10857v2cs.AI

호라이즌 불변 행동 복제를 위한 자기 지도 키프레임 발견

Self-Supervised Keyframe Discovery for Horizon-Invariant Behavior Cloning

Behavior cloning (BC) in non-Markovian environments is a challenging problem because policies have to reason over contextual information over long horizons. Existing policy architectures rely on recurrent or attention-based mechanisms to capture long-term dependencies. However, recurrent models suffer from hidden-state collapse and gradient instability under backpropagation through time while attention-based models are fundamentally limited by context length. To address these issues, we propose Keyframe Mnemonics, a novel self-supervised method that $\textit{discovers}$ a set of information-critical observations ($\textit{mnemonics}$) by learning an objective from randomly sampled past observations and using it as a reward for keyframe selection. We then train a BC policy that conditions on the discovered keyframes to model the action distribution. Under the simplicity bias assumption, our formulation provides context retention guarantees over an infinite horizon, while maintaining a small set of decision-relevant keyframes in the policy's working memory. We evaluate our method on synthetic memory domains, where mnemonic-conditioned BC policies achieve $100$% success rates (SR) and generalize to horizons orders of magnitude beyond training without performance degradation. Moreover, we evaluate on a memory-intensive robot manipulation benchmark, achieving a $13.9$% average absolute SR improvement over the strongest baseline across $23$ tasks. Finally, we demonstrate generalization with $80$% SR at $20\times$ longer time horizons on a real robot. Code and videos available at https://keyframe-mnemonics.github.io.

arXiv:2610.10600v3stat.ML

VC 학습의 최적 정보 복잡도

The optimal information complexity of VC learning

Steinke와 Zakynthinou(2020)는 알고리즘 의존적 정보이론적 양에 기반하여 학습 알고리즘의 정보 복잡도를 분석하는 조건부 상호정보량(CMI) 프레임워크를 도입했습니다. 본 연구에서는 이러한 양 중 하나인 평가된 조건부 상호정보량(eCMI)을 연구합니다. CMI를 통한 알고리즘 의존적 분석으로부터 VC 클래스에 대한 최적의 PAC 보장을 복원할 수 있는지는 흥미로운 질문이었습니다. 본 논문에서는 개념 클래스의 VC 차원이 d일 때, 실현 가능(realizable)한 경우에서 eCMI가 O(d) 차수인 학습 알고리즘을 구축함으로써 이러한 보장을 복원할 수 있음을 보입니다. 구체적으로 본 알고리즘은 최적의 기댓값 일반화 보장을 갖는 5개 기반 학습기의 무작위화된 다수결(Majority-of-5) 방식입니다.

arXiv:2610.10520v2cs.LG

그래프 기하학 증류: GNN에서 MLP로의 지식 격차

Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs

GNN-to-MLP distillation aims to retain the predictive accuracy of a message-passing teacher while deploying a graph-free MLP at inference. Existing methods mainly transfer node-wise predictions or use confidence-based reweighting, but they do not specify where the student should preserve the teacher's graph-induced geometry. We show that this omission leads to two spectral failure modes in the student's representation space. On sparse graphs, the student suffers from spectral underfit, missing high-energy teacher directions induced by aggregation. On dense graphs, it suffers from spectral overfit, retaining spurious directions that the teacher has collapsed through aggregation. Motivated by an energy-weighted teacher-student alignment objective, we propose Graph Geometry-aware MLP (G^2MLP), a training-time distillation framework guided by Ollivier-Ricci curvature. Curvature acts as a local proxy for allocating supervision between prediction-level and representation-level alignment. The deployed model is a standard MLP with no graph access at inference. Across node-classification benchmarks, G^2MLP improves over graph-free distillation baselines in most settings, reduces the teacher-student rank gap in both regimes, and transfers to Graph Transformer teachers and link prediction.

arXiv:2610.09471v5cs.LG

문맥 내 학습자는 언제 가설 공간을 확장해야 하는가?

When Should an In-Context Learner Expand Its Hypothesis Space?

Learning systems adapt quickly within a familiar family of models. The harder step is deciding, from a prediction failure, whether opening a richer family is worth its cost. We cast this as a costly sequential decision and introduce the Structural Revision Environment (SRE), which produces matched failures from noise, exceptions, change within the family or structure outside it, varies the price of expansion and the remaining horizon independently of the evidence, and admits an exact normative solution of the one-shot decision. The solution shows that revision is a value boundary, not an evidence threshold: except at the lowest prices, the richer family becomes the more probable explanation long before expanding becomes optimal. Small Transformers trained in SRE reproduce the expansion boundary from utility alone. Pretrained 7B and 8B language models fail at a definite point: their predictions carry a failure-sensitive signal that their decisions do not reflect. Told the gain of expanding, they weigh it against price and horizon only when they reason, and the one model that finishes its reasoning on the history alone still does not estimate that gain. Controlled post-training of the small Transformers moves the prior and the sharpness of predictions, and neither moves the criterion.

arXiv:2610.08446v2cs.AI

AssemState: 설명서 및 물리적 상태 기반 추론을 통한 제로샷 가구 조립

AssemState: Manual and Physical-State-Guided Reasoning for Zero-shot Furniture Assembly

멀티모달 대규모 언어 모델(MLLM)은 시각적 이해에서 상당한 발전을 이루었으나, 물리적 환경(environment)과 통합된 정밀한 3D 공간 추론은 여전히 어려운 과제로 남아 있습니다. 가구 조립은 도해 설명서로부터 단계별 작업을 복원할 뿐만 아니라, 부품이 환경 및 이전에 조립된 구성 요소와 물리적으로 상호작용할 수 있도록 의미론적 결합 관계를 6D 포즈 업데이트로 변환해야 합니다. 이러한 문제를 연구하기 위해, 본 논문에서는 설명서 및 물리적 상태를 기반으로 가구를 조립하는 제로샷(zero-shot) 프레임워크인 AssemState를 제안합니다. 이 방법은 먼저 앵커 기반 경계 조립 상태를 활용하여 설명서 페이지를 단일 부품 작업으로 분해하고 조립 트리를 복원합니다. 그런 다음, 반복적인 사후 상태 피드백 정제를 사용하여 연속적인 SE(3) 업데이트 및 수정을 유도하고, 시뮬레이션 기반 해제 테스트를 통해 물리적 타당성을 검증합니다. 실험 결과에 따르면 AssemState는 가장 강력한 이전 기준선(baseline)과 비교하여 조립 트리 복원에 대해 F1을 38.58%에서 62.80%로, 트리 정확 일치(Tree Exact Match)를 28.24%에서 53.92%로 향상시켰습니다. 독립적으로 평가된 243개의 부품 수준 작업에서 제안된 반복적 정제는 판정 승인 작업을 0에서 5.3%로 개선하고 평균 챔퍼 거리(Chamfer distance)를 5.4111에서 1.7744로 줄였습니다. 그러나 시각적으로 그럴듯한 후보 포즈라 하더라도 충돌, 공중 부양, 거울 반사 방향 오류, 불완전한 안착, 반대쪽 부착 등의 문제가 여전히 발생할 수 있습니다. 이러한 결과는 AssemState가 작업 구조 복원 및 선별된 국소 포즈 지표를 개선하는 한편, 공간 관계 추론에 있어서 MLLM의 한계가 여전히 존재함을 보여줍니다.

arXiv:2610.08443v3stat.CO

GPU-accelerated Regularized Vector Multiplicative Error Models for Positive-valued Financial Time Series

The logarithmic vector multiplicative error model (log-vMEM) provides a methodology for modeling and forecasting multivariate positive-valued time series. However, the number of autoregressive parameters increases rapidly with the system dimension and lag order, posing challenges for estimation and model selection. This paper develops a regularized estimation framework for the log-vMEM with multivariate Gamma errors by incorporating hierarchical lag structures to induce sparsity and accommodate heterogeneous lag dynamics. We consider three hierarchical lag structures: componentwise, own-other, and elementwise, in combination with group lasso, adaptive group lasso, group MCP, and group SCAD penalties. Parameter estimation is carried out through an alternating block-coordinate optimization algorithm with sequential Gauss-Seidel updates, while GPU-accelerated numerical quadrature is employed to address the computational burden of evaluating the multivariate Gamma likelihood. Simulation experiments investigate parameter recovery and lag selection under different sparsity structures. The real data analysis is carried out on the realized volatility measures for Microsoft (NASDAQ:MSFT) which allows a comparison of the competing regularization approaches in an empirical financial setting. Together, these developments provide a computational framework for sparse multivariate volatility modeling with flexible lag structures and dependent Gamma errors.

arXiv:2610.07473v2cs.AI

From Dialogue to Diagnostic Hypotheses: Evaluating LLM Agents for Psychiatric Assessment

Although LLMs show promising diagnostic reasoning on supplied clinical evidence, their ability to elicit such evidence through free-form psychiatric interviews remains less understood. We evaluate interview quality and diagnostic hypotheses using simulated patients and human participants. For controlled simulation, we introduce PsyCPG, transforming clinical cases into structured patient profiles. On 30 paired cases, experts rate PsyCPG simulations higher in quality than raw-case role-play (4.50 versus 3.83/5). We introduce PsyCIDRA, a dual-agent framework: its interviewer agent uses a notepad tool, expert-written skills, and ICD-11 reference tools to guide interviews; its diagnostic agent analyzes the transcript and produces a report for expert review, ranking differential diagnoses with supporting, conflicting, and missing evidence. Across four models and 53 development cases, PsyCIDRA achieves higher observed diagnostic agreement than direct prompting. On 81 held-out simulated cases, exact-code rank-1 accuracy reaches 60.5% versus 51.9%. We conduct 101 interviews with human participants; two psychologists, blinded to interview condition and model reports, independently assess interview quality and transcript-supported hypotheses. In separate arms, PsyCIDRA agrees with experts on whether to issue a hypothesis in 79.6% of cases versus 65.4% for direct prompting. This study provides an empirical basis for understanding and evaluating LLM agents in free-form psychiatric interviewing.

arXiv:2610.07417v2stat.ML

희소 RKHS 다양체를 통한 함수 공간에서의 베이지안 최적화

Bayesian Optimization on Function Spaces via Sparse RKHS Manifolds

베이지안 최적화(Bayesian Optimization, BO)는 벡터 입력을 갖는 블랙박스 함수를 최소화하기 위해 널리 사용되는 방법론으로 자리 잡았습니다. 그러나 이러한 매개변수(parameter) 벡터는 본질적으로 함수적 관계를 띠는 대상의 이산화(discretization)에서 기인하는 경우가 많습니다. 최근 여러 연구에서는 최적화 대상 변수가 유한 차원 벡터 공간의 원소가 아니라 무한 차원 함수 공간의 원소인 함수형 베이지안 최적화(Functional Bayesian Optimization, FBO) 설정을 다루었습니다. 본 연구에서는 커널 함수에서 희소 표현(sparse representation)을 갖는 재생 커널 힐베르트 공간(Reproducing Kernel Hilbert Space, RKHS)의 부분집합을 탐색하며 커널 위치와 해당 계수를 모두 최적화하는 단순한 FBO 접근법인 $L^0$ 다양체 최적화($L^0$ Manifold Optimization, L0MO)를 제안합니다. 우리는 제안 방법과 기존 방법 간의 관계를 상세히 논의하고, 선행 연구들을 바라볼 수 있는 통합적 관점을 제시합니다. 기존 최고 수준(SOTA) 방법 대비 제안 방법의 성능을 평가하기 위해 광범위한 계산 연구를 수행하며, 그 과정에서 표준 유한 차원 벤치마크를 무한 차원 도메인으로 변환한 새로운 벤치마크 테스트 함수 세트를 개발합니다. 실험을 통해 제안 방법이 다양한 테스트 벤치마크에 걸쳐 전반적으로 우수한 성능을 달성함을 입증합니다.

arXiv:2610.06274v2cs.LG

When the Agent Edits Its Own Limits: Evaluating a Write-Time Floor Against Prompt Prohibitions

Many agent runtimes give the agent a tool for editing its own configuration. Some settings grant abilities, such as enabling a tool. Others set the agent's limits: where it may write, who may message it, which address it listens on, how callers authenticate, and the gate that blocks risky writes. We measure how often models widen those limits while handling ordinary requests, and how well two kinds of defence hold: a prohibition in the system prompt, in three wordings, and a check at the point of writing that we call a containment floor. Every case records whether a protected write was attempted and whether it was saved, read from the tool's arguments and from the file rather than from the model's reply. With no defence, a frontier model saved a protected value on 25 of 72 requests, often when the request never named the field. A prompt that listed the protected field names stopped every request that used those names and did not reduce writes on requests that only described the goal; a prompt that described the forbidden effects did the reverse. With the floor, 0 of 167 protected writes were saved while 65 were attempted. A single-pass pilot on two further models shows the same pattern. We draw four lessons for evaluating agent safeguards: count attempts beside outcomes, test a prompt defence with requests written in a different vocabulary from the prompt, check that a zero means the safeguard fired, and log what the harness sends to the model, because a request-ordering defect in our own runtime was invisible in the transcripts and was found again, independently, by a later benchmark.

arXiv:2610.05900v4cs.CV

Sub-Voxel Displacement and Ipsilateral Territory Decoupling for 3D Bilateral Coronary Artery Tree Reconstruction in Cardiac CT Angiography

Reconstructing topologically complete, sub-voxel-accurate 3D bilateral coronary artery trees from cardiac CT angiography (CCTA) requires simultaneously recovering continuous centerline trajectories and separating dense left/right coronary territories without cross-ostial collisions. We present a five-stage tubular drift and ipsilateral territory decoupling framework (M1-M5). Stage M1 pairs a 398,843-parameter multi-scale 3D pyramid encoder with a four-head tubular field network predicting vesselness probability, signed tubular boundary distance, unit tangent vectors, and continuous 3D centripetal displacement vectors. Stage M2 introduces recall-preserving multi-task optimization with 7-group AdamW parameter decoupling, L2-SP backbone regularization, and per-step anchor shrinkage. Stage M3 calibrates cylindrical Valsalva-sinotubular-junction wall coordinates and native-resolution orifice contact expectations for robust left and right ostial rooting. Stage M4 constructs a scalar-potential-weighted minimum-cost arborescence followed by normal-plane cross-sectional pullback and centripetal trajectory pruning. Stage M5 enforces coverage-guarded ipsilateral-field-coupled bilateral collision reduction and subtree handover. Across a 14-case held-out development cohort (Dev-14, 17-arm forward incremental build and 9-arm leave-one-out ablation) and an independent 50-case out-of-sample evaluation cohort (OOS-50) benchmarked against seven external architectures (nnU-Net, 3D UX-Net, Swin-UNETR, MedNeXt, CS2-Net, clDice-UNet, and Point-UNet), M1-M5 reduces bilateral chamfer distance on Dev-14 from 3.4381 mm to 2.2759 mm and achieves 2.7671 mm on OOS-50 with 7.11% bilateral collision rate and 79.7% distal branch completeness.

arXiv:2610.05771v2cs.CV

Controllable Road Marking Generation

Lane and road markings provide critical guidance for vehicle navigation and multi-agent coordination, yet authoring them at scale remains a manual workflow that limits quantitative analysis and scenario testing. We introduce Controllable Road Marking Generation, which synthesizes a missing center-region marking layout from a drivable-area mask, optional outer-ring markings, and a textual description. Our benchmark uses deterministic, metadata-derived prompts and three output channels: lane dividers, road dividers, and pedestrian crossings. We develop a conditional bird's-eye-view (BEV) pipeline that combines (i) a text-conditioned latent rectified-flow DiT trained with a topology-aware auxiliary loss, (ii) Gaussian-blurred training targets that provide smoother supervision for thin, sparse markings, and (iii) Structured Gaussian Render (SGR), a training-free post-process that recovers crisp divider geometry by extracting polylines, fitting cubic B\'ezier curves, and re-rendering them as anisotropic super-Gaussian primitives. On 4,597 Argoverse~2 test tiles, our system achieves Buffered F1 of 76.8 and clDice of 50.2, compared with 38.8 and 24.6 for an adapted mask-refinement baseline. When trained independently on the Waymo Open Dataset, it yields 88.0 Buffered F1 and 66.2 clDice. Component ablations show complementary connectivity gains from topology-aware supervision and SGR. Text-editing experiments reveal that stronger guidance improves edit success but also increases changes to non-target structures. We see this framework as a step toward simulation-ready road-marking variation, automated map completion, and early-stage infrastructure design exploration.

arXiv:2610.05533v4cs.LG

What Does a Harness Repair? Four Preregistered Studies Around a Longevity-Tuned Language Model

A harness search around a frozen longevity-tuned 1.7B model reported a rescue. An audit traced its gain to rows the start left unparsed and found that the caller sent reference answers as assistant turns; replication through the same pipeline missed this defect. We preregistered four studies and report every endpoint. In the first two, thinking off raised test accuracy over capped thinking in 7 of 12 cells, mainly by parsing rows the cap left unparsed, and was not meaningfully worse than a rescue in 7 of 9 small-model test cells. Replication reproduced an injected defect's distortion in 9 of 20 evaluable cases. At 14B, two GEPA harnesses beat thinking off by 34 test points but were not detectably above the strongest constant label. In the third study, GEPA beat the constant on FPB (confirmatory), thinking off was not meaningfully worse than any BoolQ harness, and 3 of 4 MMLU-Pro searches kept their seed; the fourth comparison was inconclusive on test. On exploratory Autocast, harnesses were within 0.7 points of the constant and gave one label in 93.1% to 96.8% of parsed answers on test. On the prediction tasks (14B mortality, Autocast) the harnesses ended near the constant label. The fourth study varied search-set skew and objective on CREAK and resolved Manifold questions. All its results are descriptive because render hashes were not committed before the first replication call (D001). Minority-recall estimates were lower at 80:20 than at 50:50 in all four test cells. The balanced objective and re-selection from each search's pool had higher minority- and lower majority-recall estimates in all four cells. Five of their eight minority-recall intervals included zero, as did all eight balanced-accuracy contrast intervals. This does not show that skew explains the earlier collapse.

arXiv:2610.05037v2cs.AR

SparseCraft: Agentic Hardware-Software Co-Optimization for Sparse Computing

Sparse-accelerator design spaces are usually searched against analytical models, so a design is judged by a model's prediction rather than by a measurement of the hardware. SparseCraft is a closed loop on the CHIA framework in which a large language model (LLM) agent co-designs the configuration, memories and sparse-kernel schedule of a Gemmini accelerator. A candidate counts only after it passes a legality check, cycle-accurate simulation, a bit-for-bit comparison of every output and synthesis. Post-synthesis timing analysis checks its clock, and its energy is computed from measured counters, with per-event energies validated against gate-level and SRAM-macro data. We compare the agent with expert Gemmini configurations and with random, greedy and multi-objective TPE search at the agent's budget of 15 evaluations, all measured by the same flow. On a GraphChallenge sparse-DNN layer, the agent's designs reach 159-164$\times$ the inverse energy-delay-area product of stock Gemmini and 2.07-2.14$\times$ that of the best expert configuration. Their Pareto hypervolume is 0.5557-0.5704, against at most 0.2806 for search. On a power-law graph and a pruned BERT layer, the agent raises perf/W by 11.45$\times$ and 4.43$\times$, and held-out transfer shows that its designs specialise to their workload. The loop is also resilient to faulty designs. When a design fails a check, a repair agent diagnoses the failure from its evidence and fixes it automatically, or declines when the only fix would undo the intended change. In a fault-injection benchmark, it fixed every scorable fault at its first attempt.

arXiv:2610.04510v3cs.AI

ManifoldCache: Training-Free Diffusion Acceleration via Constraint Manifold Caching

Diffusion models for structured scientific generation must produce samples satisfying hard geometric constraints imposed by physics, chemistry, or biology, yet inference in these settings is prohibitively slow, demanding hundreds to thousands of neural-function evaluations per sample. We unify eight state-of-the-art models spanning medical volumetrics, molecular conformations, protein backbone design, crystal structure prediction, and multi-view 3D scenes under a single abstraction, Constraint-Manifold Diffusion Models (CMDMs), in which the target distribution is supported on a manifold defined by an externally specified constraint map. All existing acceleration families fail on this class: quantization exhausts memory on high-dimensional volumetric operators; pruning breaks constraint fidelity; fast ODE solvers allow trajectories to drift off the constraint manifold; and feature-caching heuristics are blind to constraint geometry, inducing mode confusion in the high-noise regime. We introduce ManifoldCache, the first training-free, data-free accelerator designed from first principles for CMDMs. The key insight is that the conditional score decomposes orthogonally into a normal component, which enforces constraint satisfaction, and a tangential component, which navigates within the manifold. Exploiting this structure, we prove that the noise-schedule midpoint is a sharp safe-caching boundary: caching before it incurs provably bounded error, while caching after it guarantees a strictly positive fraction of trajectories suffer mode confusion, a gap that persists up to the boundary. We further prove that deeper network blocks admit provably larger certified cache strides within the safe phase, as a consequence of the score decomposition propagating through block Jacobians. The resulting schedule requires no calibration data, along with zero training overhead.

arXiv:2610.04273v2cs.LG

Integrated Imputation-Classification for Supervised Learning with Missing Data

We study supervised classification problems with missing feature values. Existing approaches often decouple imputation from classification, producing imputations that may be plausible but uninformative for prediction. Instead, we propose the Integrated Imputation and Classification Network (IICN), which jointly trains an imputer and an ${(n{+}1)}$-class discriminator adversarially with a single class-supervised objective, where the discriminator learns to distinguish among the $n$ true classes and an additional ``imputed'' class. We prove that it admits a global optimum where the imputer and discriminator together implement marginalization over missing coordinates and yield a Bayes-optimal classifier. We evaluate IICN on FashionMNIST, CIFAR-10, and tabular datasets with naturally occurring missingness. IICN outperforms classical impute-then-classify pipelines and recent generative baselines, showing strong robustness and accuracy in challenging settings.