Papers

Large language models are increasingly expected to support diverse chemical reasoning capabilities within a unified model. One approach is to develop specialized capabilities separately and consolidate them through multi-teacher on-policy distillation, but this raises two questions: how should specialization be organized, and how should specialist guidance be integrated? We introduce ChemOPD, which addresses both. We estimate task affinities from supervised fine-tuning gradients and solve a constrained mixed-integer program(MIP) to construct partially overlapping specialist groups. During distillation, we retain a generalist teacher trained on all tasks so that specialist guidance supplements rather than replaces its supervision. Our anchor-residual objective gradually increases the routed specialist's contribution on student-generated responses. On ChemCoTBench, affinity-guided specialization produces task-dependent gains over the generalist teacher and improves several capabilities beyond semantic task grouping. Yet stronger teacher-side performance does not automatically yield stronger students: with the same specialists and routes, anchor-residual OPD improves most reported metrics over specialist-only distillation and realizes a larger share of the available teacher gains. These results highlight specialization and capability integration as connected but distinct design problems in chemical reasoning.

Construction environments evolve continuously, causing large geometric changes that degrade static mapping and registration performance. This necessitates active perception, where robots deliberately select sensing configurations to resolve the environment's current state. We present DeltaSeek, an initial framework toward active perception in evolving built environments. While our broader objective is a system that reasons about where, how, and when to observe, this paper addresses a critical prerequisite: how a robot's sensing embodiment constrains the observations it can acquire. We formalize an embodiment's permissible observation set and evaluate with a Husky A300 equipped with a UR5e on an IFC-derived benchmark under chassis-mounted and wrist-mounted RGB-D configurations, scoring observations by geometric visibility and effort by drivable distance. In a room-scale scene with eight controlled changes spanning four observability conditions, exhaustive evaluation over 240 permissible base poses and five arm postures shows that two changes admit no chassis viewpoint whatsoever, while the wrist camera resolves both. For changes observed by both embodiments, the median base travel is $6.0$~m for the wrist camera and $15.2$~m for the chassis camera. These results distinguish sensing limitations from acquisition costs, clarifying whether an observation is impossible or simply requires more travel.

Weak-to-strong generalization (W2SG) occurs when a student trained on a teacher's predictions outperforms that teacher. We study when this happens under fully converged, ridgeless two-stage learning, with no early stopping, no explicit regularization, and no assumption that the student is more expressive than the teacher. In two-stage linear regression, a teacher is fit from $n$ labeled examples and a student is trained solely on the teacher's predictions on $m$ fresh, unlabeled inputs. Although both stages share the same hypothesis class and the same training rule, we show that the student outperforms the teacher exactly when $m$ lies in an explicit intermediate range: too few pseudo-labels leave the student without enough signal, too many let it inherit the teacher's noise. Under power-law covariance, we derive this range in closed form as a function of the spectral decay and noise level, including regimes where the improving region splits into two disjoint intervals of $m$. We then study a random-feature model in which the student has strictly more features than the teacher, and identify two regimes, again given by explicit thresholds: one where improvement occurs only for $m$ in a bounded interval, and one where it occurs only once the student width $N_S$ exceeds an explicit threshold. Both regimes are governed by a single "active-bottleneck" principle: whichever of $m$ or $N_S$ is scarcer controls how much teacher error is filtered out, while increasing the other resource only reduces estimation noise. Together, these results show that finite data and finite width can themselves regularize a two-stage learner, with no explicit mechanism doing so.

Diffusion models are now widely used in Bayesian inverse problems in imaging as priors, where latent diffusion models are often used for larger scale problems to keep the computational complexity and model-size manageable. Unfortunately, the auto-encoder based compression results in loss of spatial detail. In addition, the optimization is converted to a non-linear problem. In this paper, we introduce a posterior sampling algorithm customized for the pyramidal/cascaded architecture, which relies on a coarse to fine hierarchical strategy to generate images in the pixel domain. We present CLIMB-Flow which alternates between three steps: an end-point estimation from the current coarse and noisy image, data-consistent update of the clean image, and re-noising it back to the level the network expects. Together these steps sample the posterior at that scale using an approximate Gibbs sampling from two conditional distributions. Experiments on ImageNet, CelebA, AFHQ and fastMRI span inpainting, deblurring, super-resolution and accelerated MRI, with PSNR gains of 1.37-7.66 dB over the strongest competing method on CelebA and pixel-domain reconstruction up to 512x512.

Adversarial attacks have long posed a fundamental threat to machine learning systems. As multimodal large language models (MLLMs) rapidly evolve and become widely deployed, assessing their vulnerability to such attacks is essential for their safe use. In this work, we investigate whether a single adversarial image can consistently mislead diverse frontier MLLMs in black-box settings. We propose O-Attack, a highly transferable black-box attack framework. This framework builds on our insight that surrogate models contain a broad, high-level, cross-modally aligned semantic space. This space extends beyond final-layer outputs and provides multiple semantically consistent representations that remain underexploited by existing attacks. Within this space, O-Attack anchors aligned representations, progressively broadens semantic conditions, and optimizes perturbations through semantic consensus to promote consistent target alignment. By fully exploiting this space with the same surrogate models as M-Attack, O-Attack raises attack success rates on GPT-5.4 (29.1% to 77.2%), Claude-4.6 (42.8% to 81.6%), and Gemini-3.1 (38.2% to 80.9%). Extensive experiments across 24 MLLMs show that O-Attack outperforms six state-of-the-art methods in black-box transferability, with consistent effectiveness across prompts and improved efficiency and imperceptibility. This work exposes the practical safety risks posed by black-box adversarial attacks against frontier MLLMs, underscoring the need for more rigorous robustness evaluation and more effective defenses.

World-action models can jointly predict future visual observations and robot actions. However, discrepancies may exist between their visual predictions and the consequences implied by generated actions. We observe that WAMs can often generate visually plausible task-completion outcomes before producing action sequences that reliably achieve them. Consequently, we treat the WAM-generated visual prediction as a goal-conditioned visual proposal rather than a directly executable plan. We use a frozen action-conditioned world model to predict action-conditioned consequences and construct feedback based on consistency between the two future predictions and alignment with the terminal goal. Leveraging this feedback, we employ Flow Policy Optimization (FPO) to optimize the action head of the WAM. This framework avoids online robot interaction and additional training of task-specific reward models. Across four real-world UR5 manipulation tasks, our method increases the mean success rate from 43.4% to 75.1%, compared with 61.4% for $\pi_{0.5}$. These results show that cross-model prediction discrepancy can provide useful feedback for improving robot policies under the evaluated manipulation tasks. Website: https://imagine-to-achieve.github.io/

Tor is a widely used anonymity network that provides network privacy by routing client communications through a sequence of relays to their destinations. An adversary observing a single Tor relay cannot readily link clients to their destinations, as doing so requires identifying the other relays along the client circuit. Identifying these relays from traffic observed at the monitored relay alone is challenging: each relay carries traffic for many circuits simultaneously, and client circuits typically last only about 10 min, limiting observations of a given circuit. In contrast, Tor onion services use introduction circuits that remain fixed for 18 to 24 h. We develop an intersection attack that exploits this design, allowing an adaptive adversary observing one Tor relay at a time to identify the network location of an onion service. We evaluate the attack on the live Tor network against a self-operated onion service carrying genuine third-party traffic. Across nine end-to-end experiments, the attack reconstructs the complete Tor circuit in every run, with an estimated median time of 2.2 h. To mitigate this attack, we propose a mechanism that periodically rebuilds internal introduction circuits to limit evidence accumulation without changing publicly advertised information.

Guessing Random Additive Noise Decoding-Aided Macrosymbol (GRAND-AM) provides a coding-based solution for non-orthogonal multiple access (NOMA) systems, enabling joint multiuser detection and error correction. GRAND-AM outperforms prior methods in symbol error rates while providing hard-detection output. We introduce soft-output GRAND-AM (SOGRAND-AM), a macrosymbol-level joint multiuser detection and decoding framework which generates calibrated bitwise a posteriori error probabilities. SOGRAND-AM extends GRAND-AM by enabling soft-decision outer forward error correction decoders, transparently integrating with existing decoders. SOGRAND-AM achieves approximately 2 dB improvement in bit error rate at the output of the outer soft-input decoder compared to hard input with GRAND-AM.

Optimization layers enable the incorporation of structured constraints and decision problems into learning systems. Training such systems requires differentiating through the embedded optimization problem, which can be challenging for general conic programs. We introduce dOPT, a solver-agnostic framework that, rather than differentiating the full conic formulation, reduces it at a computed primal-dual solution to an equality-constrained quadratic program that preserves the reference solution and its first-order sensitivity. The reduction captures the local first- and second-order conic geometry relevant to differentiation and remains well defined at singular configurations. Computing solution derivatives then requires a single symmetric linear solve, independently of the forward solver. We derive explicit reductions for convex NLPs, QPs, SOCPs, and SDPs. Numerical experiments validate the computed gradients and show favorable backward-pass scalability, with substantial speedups over existing differentiable conic optimization methods as problem size increases.

An agent harness controls how a language model accesses information, uses tools, preserves memory, and checks its work. Improving this software is costly when each evaluation requires a long interaction with an environment. We introduce Vestrum, a framework that turns failures in execution traces into scoped harness changes without training the task model. Its organizing overhypothesis is that tasks of a shared kind may exhibit recurring failures whose remedies transfer within that kind. Vestrum expresses failures as recognizable classes, proposes changes across verification, retrieval, decomposition, and knowledge synthesis, and screens their scope before evaluating them as a bundle. A persistent lessons file informs subsequent proposals. Across five settings and two baseline harnesses, the frozen harnesses improve held-out performance: UltraHorizon rises from 47.6 to 59.8 over GAM, Terminal-Bench 4 Hard from 63.7% to 70.3% of checks passed over Claude Code on eight held-out tasks at 1.03x test cost, and cell-type annotation agreement from 67.5% to 77.8% on held-out sections of one slide, alongside gains on LoCoMo and AMA-Bench. Across our searches, verification grounded in evidence helped both intermediate steps and final answers, at lower cost at intermediate steps, while critics asked to rebuild finished answers broke more than they repaired. On the three memory benchmarks, Vestrum also scores above the evaluated GEPA configurations in every paired evaluation.

Visual anomaly detectors identify deviations from known-normal data, but their anomaly signals may mix evidence of actual anomalies with benign visual variation. We investigate whether automated interpretability can augment visual anomaly detectors by identifying and intervening on different components of this signal. We decompose PatchCore nearest-normal residuals into sparse features using Sparse Autoencoders (SAEs), and provide high-activation and contrastive non-active examples to a Multimodal LLM, which describes each feature and labels it as anomaly, distractor, or uncertain. These labels guide interventions in the SAE hidden representation, where distractor features are suppressed and anomaly features amplified. The edited representation is then used to reconstruct patch embeddings, which are rescored with PatchCore. Across 40 categories from four benchmarks, applying both interventions jointly improves macro-average image-level AUROC from 0.8724 to 0.8857 on source data and from 0.8066 to 0.8210 under synthetic corruptions. On three additional RobustAD categories with real acquisition shifts, the same interventions improve AUROC from 0.8745 to 0.9056 on source data and from 0.6069 to 0.6599 under real acquisition shifts. Finally, individual feature interventions across all 43 categories show that the MLLM labels are aligned in aggregate with how features differently affect normal and anomalous images.

On-policy distillation (OPD) bridges teacher supervision and student behavior, but different teacher-student tokenizers introduce misalignment in both input tokenization (#1) and output logits (#2). Existing approaches address the former by matching same-text spans or converting tokens to bytes, often losing fine-grained token information or disrupting the native-token paradigm, while for the latter, strategies such as ranking, padding, or key-token selection retain only shared logit dimensions, resulting in much distribution loss. In this paper, we propose Dual-Vocabulary Language Model (DVLM), which replaces the teacher's LM head with a new student-vocabulary projection head and obtains full-dimensional student logits (for #2). To support student tokens (for #1), it takes a Parallel-Tokenized Sequence (PTS) as input, which concatenates the original teacher-tokenized sequence and a re-tokenized sequence formed by independently converting each student token into a teacher-token group. To avoid inference inconsistency with the original teacher tokens, the Hybrid-Prefix Attention (HPA) further restricts re-tokenized groups to their corresponding teacher prefix and uses its last state as the aggregation of the original student-token representation for projection into the student vocabulary space. Similarly, via the combined use of PTS and HPA, the DVLM teacher can provide distribution-aligned supervision with the student's input-tokenization and output-logit during OPD. Experimental results demonstrate that our DVLM teacher has a similar converged loss as the original teacher model and enables student models to improve performance across six reasoning tasks.

We characterize Bregman Douglas-Rachford splitting (BDRS) for unregularized discrete optimal transport and develop an anytime primal-dual certificate in linear memory. We first establish that BDRS coincides with warm-started Inexact Proximal point method for exact Optimal Transport (IPOT) using a single inner Sinkhorn iteration. By eliminating the primal transport plan from the updates, we derive an equivalent dual formulation that reveals BDRS as annealed Sinkhorn under an implicit inverse-linear temperature schedule, with an additional log-scaling momentum term and a cooler kernel. While this explains the role of the temperature parameter in BDRS as an initial temperature, it also reduces the solver's memory requirement from quadratic to linear. Utilizing this annealing perspective, we introduce overrelaxed BDRS, which combines annealing and overrelaxed scaling within a single recursion. We derive a primal-dual certificate for both methods that can be evaluated in linear memory without transport plan construction, thus providing a computable stopping rule. On the DOTmark benchmark, combining momentum with the cooler kernel produces substantially smaller optimality gaps than annealed Sinkhorn under the same schedule. For pixel-level color transfer between $1024\times1024$ images, BDRS attains a lower repaired transport cost than MDOT-TNT with a 9$\times$ speed up, reaching a relative duality gap of $1.59\%$ in 24 minutes. We further demonstrate a color transfer with $4238\times2365$ images, yielding 10 million pixels per image and approximately one hundred trillion implicit transport entries, reaching a best relative duality gap of $2.41\%$ and $2.80\%$ within 35 hours in each direction on a single NVIDIA L40S GPU.

The coroutine has broadly diffused in the practice of concurrent programming in the form of generators and asynchronous functions as well as processes communicating through pipes. We wanted to use coroutines in Python to create single-threaded Unix-style pipelines. Unfortunately, available solutions in Python are cumbersome to use. The JavaScript push-stream protocol appeared to be a good alternative. However, its specification is incomplete and ambiguous. We have used TLA$^{+}$ and the TLC model checker to re-derive the protocol and obtain protocol-specific verification tools. In this paper, we present a formal specification of a push-stream protocol that 1) seamlessly combines synchronous and asynchronous modules, encapsulating the choice within each module; 2) provides flow control without using bounded buffers; 3) gracefully and unambiguously terminates; 4) does not require dynamic allocation of objects on the heap. In addition to completely describing expected behaviours, our specification improves on the original design by 1) allowing the input and output of intermediate pipeline modules to terminate independently and 2) explicitly reporting when a module is pending on the execution environment, to avoid incorrect resuming. We specify the protocol as a sequence of refinement steps and derive by equivalence a specification of what an abstract module may do. We then refine the latter into a module checker that can verify concrete module specifications for conformity. We have verified the key properties of all specifications and the validity of refinement steps with TLC. In supplemental material, we provide all TLA$^{+}$ specifications, show that the protocol is sufficiently expressive to implement a superset of all original JavaScript modules, as well as a performance comparison with Python alternatives and Unix pipes.

Repeated runs of the same coding agent are known to give different benchmark scores. We ask what that variation means for a team running an agent on its own task, by intensive replication on one machine-learning task: an agent improves the training code of an XGBoost classifier for airline delays, and a holdout it never sees scores the result. Across 584 runs, we compare six agents on six open-weight model endpoints, run six agent-model pairings 52 times each under fixed settings, and repeat three of them on a larger model from the same family. Identical runs of one pairing varied more than the pairings differed from one another, so comparisons of a few runs ranked them unreliably; resolving the agent differences we observed would take tens to more than a hundred runs of each. Runs on the larger model scored clearly higher, but by less than one run-to-run standard deviation, and the gap was more than twice as large with one agent as with the others. Fewer than one run in twenty broke the task's data rules, but those runs held the highest scores. Rejecting those runs first and keeping the best compliant result among a few attempts reliably improved the delivered model, even though a few runs could not rank the agents. On flights from a later year, the delivered models kept only a third of their gain over the starting code. At list prices, cost differed more than twentyfold between two agents on the same model, mostly through the prompt cache. Agents and models should be evaluated as pairings, over repeated attempts, with compliance reported beside quality. Data, code and every delivered program: https://github.com/earino/identical-runs-different-results

Motorized two-wheelers (MTW) dominate Indian roads but remain underrepresented in driver behavior research. This study presents the first large-scale analysis of MTW driver gaze behavior in naturalistic, heterogeneous urban traffic, using the myEye2Wheeler dataset. A semantic segmentation pipeline (YOLOv11 + SAM2) was used to extract object-level gaze metrics under two attention modes: direct gaze (foveal overlap) and central vision (parafoveal monitoring). Results reveal a functional division: central vision supports broad monitoring, while direct gaze enables brief, selective sampling. Novice riders exhibit road-anchored scanning, returning to the road between object fixations, while experienced riders form longer chains of attention across multiple objects. The findings suggest that experience primarily refines temporal rhythm rather than altering allocation strategy and reduces object-class effects in gaze patterns. These findings offer new insight into MTW attention structures and inform future work on behavior modeling and safety systems.

Autoregressive text-to-speech (TTS) systems synthesize natural speech but, once trained, offer little control over speaking rate. We show that speaking rate can be steered at inference time, without retraining, by clamping a single decoder block's activation along a discovered speed axis. A decoder-block analysis recovers the rate axis, a neutral operating point, and a per-step intensity scale; at inference, the activation's projection onto this axis is set to a fixed scalar. Learning this direction from synthetically time-stretched and time-compressed speech yields rate control that largely preserves speaker identity, generalizes across model architectures, and maintains high naturalness in objective and human evaluations. Unlike standard additive steering, which breaks at the slow extreme, clamping remains stable on all three systems tested; at moderate targets, the better rule depends on the model. Finally, we show that rate information is decodable across layers but causally steerable only within a mid-depth window, and demonstrate the effectiveness of our approach on the public Seed-TTS-Eval benchmark.

How well can general-purpose multimodal models turn visual understanding and reasoning into embodied manipulation via executable code? We introduce CodeActionBench, a benchmark of 25 manipulation tasks that evaluates this capability through agentic Code-as-Policy. Without task-specific fine-tuning, demonstrations, external specialist perception or grasp modules, privileged scene state, or predefined task policies, agents should select visual evidence, form task-relevant 3D estimates, construct manipulation targets, and iteratively execute and revise their policies. A shared robot API provides RGB observations, calibrated geometric operations, robot feedback, and bounded motion, leaving task-dependent decisions to the evaluated agent. Fixed task instances, resource budgets, and a hidden physical-outcome verifier support controlled comparisons across models and harness configurations. Extensive evaluations across nine configurations and 675 attempts achieve success rates ranging from 2.7% to 73.3%. The strongest configuration, GPT-6 Astra with Codex CLI, solves 22 of 25 tasks at least once in three attempts, demonstrating the best performance while still leaving substantial room for improvement. Trajectory analyses reveal difficulties in spatial alignment, object retention, and completion judgment, including task failures despite successfully completed motions. CodeActionBench provides a controlled testbed for measuring how general-purpose models translate their capabilities into manipulation behavior and for examining typical failure scenarios in that process.

Modeling spatiotemporal coupling is a key challenge in building physical intelligence across scales, from microscopic to macroscopic. Existing models capture such structure broadly through physics-motivated dynamical formulations or learning-motivated architectures. The former provide stronger priors but may constrain flexibility, whereas the latter are more flexible but leave the spatiotemporal coupling largely implicit. We therefore seek an approach that combines flexible learning with an explicit geometric bias for jointly modeling time and space. To this end, we propose Minkowski Positional Encoding (MinkowskiPE), which uses joint temporal and spatial coordinates to parameterize Lorentz transformations applied to query and key features. With MinkowskiPE, the query-key attention score depends on position only through the relative spacetime displacement between the two tokens and is therefore invariant to global translation of the coordinates. This paradigm retains the standard dot-product attention interface and remains compatible with efficient attention implementations. We evaluate MinkowskiPE on microscopic molecular dynamics and macroscopic video prediction tasks, achieving the best results on all nine multi-trajectory molecular evaluations and reducing KTH video-prediction MSE by 9.9% relative to the best baseline while using roughly one-tenth as many parameters.

Reward models underpin the alignment of large language models, yet the dominant designs reduce each prompt–response pair to a point estimate or to a distribution from a fixed parametric family. This is at odds with human preference, which is inherently multimodal: the same response can be reasonably judged in many ways, and no single family covers all of them. To better fit this structure, we introduce DRM, a Diffusion Reward Model that recasts reward modeling as conditional density estimation over $p(\mathbf{r}\mid x,y)$. Conditioned on a frozen LLM encoder, a lightweight Diffusion Transformer denoises Gaussian noise into a reward vector, placing no parametric assumption on the output distribution and naturally representing its multimodal structure. A single architecture handles both multi-attribute regression and pairwise preference data, and at inference $N$ samples form an empirical reward distribution that can be aggregated into a scalar, a variance, or quantiles. Across five benchmarks, DRM matches or surpasses baselines under matched data and backbone, stays competitive with much larger discriminative, distributional, and generative RMs despite its modest training scale, and recovers multimodal reward structure where conventional heads collapse to a point. Uncertainty-aware rejection and lower-confidence-bound (LCB) aggregation further demonstrate that DRM can exploit distributional information beyond a scalar reward to improve reward-model decisions. Downstream RLHF experiments additionally show that using DRM as the training-time reward leads to improved policy performance, directly validating the practical benefit of diffusion-based reward modeling for RLHF training.

Understanding quantum phases of matter has long relied on physicists' intuition and mathematical tools such as symmetry and topology. Remarkably successful as these approaches have been, they provide no universal way to explore a Hamiltonian space whose organizing principle is not known in advance. In this work, we introduce a fully autonomous system combining differentiable programming and unsupervised learning for quantum phase discovery. The search evaluates ground-state data along an adaptive trajectory rather than on a predetermined parameter grid. We demonstrate the system with three different solvers and benchmark it against random sampling at equal ground-state-evaluation budgets. On a generalized cluster chain hosting up to $200$ distinct phases, the search finds up to $25$ more phases at the same budget, and matches random sampling given thirty times its budget. On a $50$-parameter Chern insulator, it reaches sectors not obtained by the simple harmonic constructions considered here, in a family whose inverse problem remains open, while recovering all sectors found by sampling. Our results establish autonomous, gradient-driven exploration of Hamiltonian space as a practical route to discovering quantum phases without phase labels or a prescribed target phase.

Multimodal learning beyond two modalities commonly leverages a specific modality (e.g., text) to bind other modalities. However, how to establish a more balanced representation space that approximates shared semantics while respecting the holistic geometry of $n$-modal data remains challenging. In this work, we present BaryBind, which aims to transport the specific modality towards the Wasserstein barycenter (WB) optimized across all modalities and introduces a volumetric alignment objective to establish a unified semantic space around the WB embedding. Specifically, we project specific modalities to the WB, which minimizes the average Wasserstein distances to multimodal distributions and serves as the anchor for subsequent alignment. We then construct a barycenter simplex, whose volume is taken as a similarity metric for global alignment centered at the WB. Experiments show that BaryBind achieves competitive performance in text-video-audio retrieval, classification, videoQA, and cross-modal generation tasks, along with robustness under modality absence and scalability to more than three modalities. Code is released at https://github.com/xl-tang3/BaryBind.

Translating natural-language access-control requirements into policies requires careful reasoning about permissions, constraints, and exceptions, and even frontier LLMs often produce policies that violate the intended authorization semantics. We construct CedarInstruct, to our knowledge the first dataset that supports both training and semantic evaluation for formally verifiable Cedar policy synthesis. It contains 5,800 scenarios across 44 domains and 1,408 representing a single synthetic organization, each with a verified target policy and an executable verification plan. On this data we introduce RAISE, which trains policy synthesizers from formal verification in two stages, verified supervised fine-tuning (SFT) followed by a reinforcement learning (RL) stage that learns from verifier signal. We find that SFT succeeds largely by letting models express authorization logic they already have, since untrained models rarely write valid Cedar but often reason correctly when they do. After SFT, how the verifier's information is used matters more than how much of it is used. Of six RL instantiations that consume progressively richer verifier signal, only RAISE-OC improves meaningfully on SFT; it turns failed checks and symbolic counterexamples into guided exploration and learns from the result with off-context GRPO. With about 5.4K verified scenarios and LoRA fine-tuning, RAISE-OC trains Qwen3.5-9B to surpass zero-shot GPT-6 Astra and Claude Opus 5 by 13.33 and 16.26 percentage points in semantic success on held-out scenarios, and training transfers to the independently constructed CedarBench.

Hydrogen-diesel dual-fuel (H2DF) combustion engines offer a promising pathway for decarbonising hard-to-electrify transport sectors, yet their highly nonlinear dynamics and coupled process variables demand constraint-aware control strategies. Model Predictive Control (MPC) meets these requirements but requires an online optimisation in every combustion cycle, which limits deployment on low-cost embedded hardware. This paper trains a feedforward deep neural network (DNN) by behaviour cloning (BC) to imitate an MPC expert, using 86,000 engine cycles of demonstration data collected at 1500 min-1 on a modified Cummins QSB 4.5-litre hydrogen dual-fuel engine. Two variants, one with process feedback and one without, are validated experimentally. Both track unseen fast-transient load steps (3-8 bar indicated mean effective pressure, IMEP) with normalised root mean square error (NRMSE) values of 7.80% and 9.03% against the expert's 8.01%, while keeping mean NOx and particulate matter emissions at or below those of the expert. Inference takes 2 ms or less on a Raspberry Pi 400 (ARM Cortex-A72 at 2.2 GHz), including 1 ms communication latency, compared to up to 7 ms for the MPC expert, a 3.5x speedup. Open-loop profiling on a low-cost ESP32 microcontroller at 180 MHz gives 4.3 ms per inference, 4x faster than required for the 18 ms cycle window. Beyond the training range the cloned policy saturates its controls but exceeds the pressure-rise-rate limit. To the authors' knowledge, this is the first experimental BC controller for cycle-to-cycle combustion control of an internal combustion engine (ICE), trained from demonstrations recorded on the engine itself.

Since the advent of knowledge distillation, KL divergence has been the standard loss in distillation. Recently, on-policy distillation (OPD) has emerged as an efficient post-training paradigm for LLMs. As a distillation method, OPD naturally inherits KL divergence as its standard loss. However, in this work, we find that KL divergence may not be necessary for OPD. We show that simply preserving the update direction is sufficient for effective OPD. As long as the update direction is toward the teacher, OPD works. More precisely, it is not the direction of every token, but the direction of a small subset of tokens where the teacher and student disagree strongly. We first show that simply assigning a reward of (+1) to tokens where the teacher probability is higher than the student probability and (-1) where it is lower, which merely encourages updates toward the teacher, reproduces almost the same training mode as OPD with reverse KL. We further show that only the direction of a small subset of tokens with large teacher-student disagreement is critical, and training works as long as their update direction is toward the teacher, even if other tokens are pulled away from the teacher. And as an application of these findings, we introduce Consensus Multi-Teacher On-Policy Distillation (C-MOPD) to improve Multi-Teacher On-Policy Distillation (MOPD). Unlike MOPD, which routes each sample to a single teacher and may cause capability conflicts across domains, C-MOPD lets every sample be supervised by all teachers. Experiments show that C-MOPD consistently outperforms MOPD on both math and code benchmarks. Our code is available at https://github.com/LeapLabTHU/KL-Free-OPD.

Conventional reflectance photoplethysmography (PPG) typically reduces tissue optical responses to one or a few detector channels. Matrix Pinhole Image Sensing (MAPIS) instead simultaneously samples multiple aperture-dependent optical signals. We analyzed 44 recordings of 30 seconds each from a single MAPIS device to characterize the spatial relationships among DC intensity, pulsatile AC, perfusion index (PI), and temporal coherence across Red and infrared channels. AC increased sublinearly with DC at both wavelengths, resulting in higher PI toward lower-DC apertures, consistent with a first-order relative path-sensitivity interpretation. A second counterintuitive spatial trend was observed in temporal coherence: lower-DC and lower-AC apertures exhibited higher autocorrelation and greater cardiac-periodic spectral concentration. IR signals also showed consistently higher temporal coherence than corresponding Red signals. These observations demonstrate that multi-aperture PPG resolves reproducible spatial relationships among optical intensity, fractional pulsatile sensitivity, and waveform coherence that are largely averaged together in conventional single-channel PPG. The present study establishes within-device relationships and motivates controlled studies of the underlying photon-path mechanisms.

As LLM-generated text becomes increasingly human-like, accurately localizing LLM-authored spans in human-LLM co-authored documents is important for attribution and accountability in cases involving copyright infringement, fraud, and other harmful uses of AI-generated content. Sentence-level detectors provide local authorship evidence, but content variation can cause score fluctuations even among sentences from the same source, creating spurious boundaries. Recovering a coherent document partition therefore remains challenging when both the number and locations of authorship transitions are unknown. We propose Local-Evidence-Aware Change-Point Detection (LA-CPD), a structured method that transforms noisy sentence-level score sequences into coherent authorship segments. Given scores from a frozen local detector, LA-CPD combines a length-weighted within-segment residual with a windowed two-mean contrast to capture segment consistency and sustained changes around candidate cut points. Dynamic programming optimizes cut locations for each candidate count, while an AIC-style criterion selects the final partition, yielding sentence labels, authorship boundaries, and maximal LLM-authored spans. On a held-out human-LLM co-authored test set, LA-CPD outperforms WCP+AIC, increasing sentence-level accuracy from 0.747 to 0.796 while improving boundary localization and LLM-span delineation.

A diffusion model can predict a follow-up medical scan from a baseline, but a clinician needs a per-voxel map of where that prediction can be trusted. Many such maps approximate the diagonal of the Tweedie posterior covariance, and are evaluated against another approximation of it, so whether the estimator or the target limits them is unclear. We compute the exact diagonal on six checkpoints across fourteen model-corpus conditions. Hutchinson at M=200 tracks it at rank agreement of at least 0.92 everywhere, yet in four of the fourteen the exact diagonal is anti-correlated with the denoising error, reaching -0.13, so a faithful estimator reproduces that reversal. All four are real-image conditions; on the models' own samples the reversal does not appear, so evaluating on generated samples flatters this family. What limits these maps is the target, not the estimator. We then introduce Tweedie Probe-Tangent (T-PT), a gradient-free residual probe that corrupts one model-supported prediction repeatedly and measures the voxel-wise variance of the denoiser's response. T-PT reads a different functional of the same Jacobian, and its exact second-order form ranks with the diagonal wherever the diagonal reverses; at thirty probes it returns a map too unstable to reproduce that ranking, while Hutchinson at M=5 already reproduces it, so T-PT there is not evidence against the reversal. We offer it as an instrument, not a better approximation. On brain MRI at full resolution, where every Jacobian-based estimator we test runs out of memory, T-PT leads a twenty-chain Monte-Carlo ensemble on five of eight endpoints inside tissue and trails it on none, at 16x fewer network evaluations; over the whole volume the ensemble leads, and fifty chains close the tissue gap. On lung CT the ensemble is ahead throughout. Both lose most of their discrimination where the change is, which remains open.

Reinforcement learning with verifiable rewards (RLVR) enhances reasoning in large language models (LLMs) through outcome-level feedback, yet recent approaches to finer-grained credit assignment often require auxiliary models, additional sampling, or privileged information. Although policy entropy provides a readily available signal, prioritizing uncertain positions under both reinforcement and penalization concentrates penalties where failed responses still retain alternatives for recovery, which can suppress opportunities for exploration. To address this, we introduce Entropic Advantage Policy Optimization (EAPO), an entropy-guided credit assignment method that treats success and failure asymmetrically. Specifically, motivated by the observation that success under uncertainty is less repeatable while confident failures tend to recur, EAPO couples normalized policy entropy with the sign of the response advantage to reinforce surprising success and correct repeated failure. It assigns stronger reinforcement to high-entropy decisions in successful responses and stronger penalties to low-entropy decisions in failed responses, while attenuating penalties at uncertain positions to preserve opportunities for recovery. By redistributing the response advantage across tokens, EAPO derives token-level credit directly from existing rollout signals without additional supervision. We validate EAPO on a range of reasoning tasks across both base and reasoning backbones, demonstrating that it achieves the best overall performance. We further show that EAPO promotes more effective exploration, broadening problem coverage and generating more diverse candidate answers.

Verified solutions are not equally useful for preparing reasoning models for reinforcement learning (RL). We present a comprehensive study of route diversity, the variation in the sequences of reasoning steps in supervised fine-tuning (SFT) data, and propose a lightweight, rule-based fingerprint to select for it. From one pool at one budget, with matched training recipes and checkpoints, selecting diverse rather than similar routes improves post-RL problem coverage across puzzles and mathematics, including on problems harder than those seen in either training stage. In synthetic experiments, route-diverse SFT improves OLMo3-7B's pass@8 by 16.9 points on environments held out from SFT. In a single-model condition, where one model writes every candidate, diverse selection gains up to 6.2 points of mean pass@8 across 10 mathematics benchmarks. Pre-RL diagnostics suggest why: diverse SFT can produce both successful and failed attempts on more prompts despite slightly lower mean accuracy, giving group-relative RL more prompts with a learning signal. On 3 open-source corpora, our CPU-only selector, without model calls, outperforms more expensive alternatives in every comparison of mean post-RL performance. These results identify reasoning-route diversity as a practical criterion for selecting SFT data that better prepares models for RL.