每天知道
你的领域
出了什么

Agent models are frequently evaluated one decision at a time, where the model predicts the next action based on the gold interaction history, which is scored against a reference. We investigate whether improvement under this protocol is predictive of improved autonomous workflow execution. We study pre-SFT and supervised fine-tuned (SFT) Qwen3 models at 4B and 14B parameters and Gemma 3 models at 4B and 12B parameters on multi-turn customer-support workflows. We find that SFT consistently improves text-turn success, and that overall next-turn success increases for every model under gold-history evaluation. However, these improvements do not transfer to autonomous workflow execution. Tool-specific gains also vary across metrics and models. None of the four SFT models succeeds under holistic workflow evaluation, with strict trajectory completion reaching at most 10.4% workflow success. Our results show that next-turn evaluation is not a reliable proxy for workflow success, motivating separate reporting of text quality, local action correctness, tool execution, and end-to-end task completion.

LLM-based multi-agent systems generate collaboration traces that record how agents plan tasks, verify intermediate results, and repair failures. Reusing these procedures requires preserving an action's prerequisites and the outputs needed by subsequent agents. Our empirical studies show that grouping these dependencies into functional memory units improves their retention, while connecting units increases retrieval of the units and links jointly required by a task. The preferred combination of units also changes between instructions and checklists, even when each combination's content is fixed across formats. Updating choices from the outcomes of each combination and format pairing outperforms scoring combinations and formats separately. These findings motivate MACE, a memory-agent co-evolution framework that adapts memory organization and agent memory use through execution feedback. Its MemGoG structure represents functional units as subgraphs of related conditions, actions, and outputs, connecting them through support, conflict, and repair relations. MACE Loop selects task-relevant units and relations within a memory budget and provides each agent with instructions or checklists for its current operation. It records the selected units, presentation formats, agent outputs, and task outcomes to update unit scores and relations for retrieval and inform subsequent presentation choices. Across eight benchmarks, MACE outperforms ten baselines with an average score of 81.11%, compared with 78.97% for the strongest baseline, SAGE.

Nearly 90% of drugs entering clinical development ultimately fail, despite billions of dollars in investment. Pharmaceutical companies therefore rely on clinical development planning (CDP) and probability of technical and regulatory success assessment to anticipate development risks, yet these decisions remain labor-intensive and subjective, requiring experts across clinical science, statistics, regulatory affairs, and competitive intelligence to jointly acquire, synthesize, and reason over heterogeneous evidence. Here, we introduce TrialAtlas, a memory-augmented multi-agent research organization for CDP that mirrors this collaborative process by coordinating specialized agents for literature synthesis, competitive trial intelligence, regulatory precedent analysis, and integrated reasoning over trial design and development risk. TrialAtlas further learns from historical clinical trials and regulatory outcomes, including prior New Drug Applications (NDAs), to ground its decisions in accumulated development experience. To evaluate these capabilities in an authentic regulatory setting, we introduce TrialAtlasBench, constructed from 291 FDA Complete Response Letters and spanning three practical tasks: detecting trial design deficiencies, recommending actionable design improvements, and predicting technical and regulatory success. TrialAtlas achieves an F1 score of 50.0% for deficiency detection, outperforming the strongest baseline by 6.1 points, and reaches 85.3% balanced accuracy and 84.7% F1 for prediction of technical and regulatory success, improving over the best baselines by 6.7 points in balanced accuracy and 12.0 points in Cohen's kappa. In expert evaluation, 86.4% of TrialAtlas-generated concerns were judged valid, compared with 83.1% for OpenAI DeepResearch and 59.3% for Gemini DeepResearch.

Text-guided video editing modifies target content while preserving the appearance and temporal coherence of unedited regions. Training-based approaches provide strong control but demand substantial data and computation. Training-free methods fall into inversion-free and inversion-based paradigms. Inversion-free approaches avoid trajectory recovery, but their source-preserving guidance can limit editing strength and leave semantic changes incomplete. Inversion-based approaches recover a latent trajectory before regeneration, where approximation errors can accumulate and cause source-content drift and temporal inconsistency. We introduce Edit-VAR, the first training-free and inversion-free framework for text-guided video editing with a pretrained visual autoregressive video model. Edit-VAR directly encodes the source video into multi-scale discrete tokens and performs probability-guided conditional token replacement for source preservation. Attention-guided token-wise and scale-aware modulation selectively relaxes source constraints over edit-relevant positions and generation stages. Scale-Decoupled Generation, implemented as late-scale constraint release, regenerates motion-consistent details and reduces texture fragmentation. Residual-guided token pruning further exploits redundancy at the final two high-resolution scales to reduce inference cost. Extensive experiments and a blind user study demonstrate that Edit-VAR outperforms existing training-free video editing methods overall in editing fidelity, source preservation, temporal coherence, and inference efficiency.

Gating the value pathway of attention reportedly improves language model pretraining, and prior studies disagree on why. We argue and provide experimental evidence that such gates supply two different things that softmax attention lacks: abstention and noise filtering. The first is abstention, which allows an attention head to output nothing, bypassing the requirement that attention weights must sum to one. The second is noise filtering, which allows the value pathway of an attention head to suppress interference from superposed features in the residual stream. In our experiments in matched models from 10M to 350M parameters, we supply abstention through a learned per-head sink logit in the softmax and noise filtering through a gate on each value. We report three empirical findings. First, the benefit of abstention, measured as the reduction in validation loss relative to a matched baseline, declines as models grow, whereas the benefit of noise filtering increases with scale. In particular, abstention accounts for nearly all of the gain from gating at 10M and filtering for most of it at 350M. Second, the best model at every scale is the one with both primitives built in. Third, injecting controlled interference into the values a head reads confirms that the gate removes such interference, and reveals that each of the two gate forms we study has a characteristic blind spot. Supplying both primitives adds negligible parameters and remains compatible with the key-value cache.

Production LLM agents are evaluated repeatedly as they evolve, but full agent benchmarks are costly to rerun. We study efficient recurring evaluation for a production analytics agent serving tens of thousands of monthly active users and report first-hand deployment experience. Using 574 historical runs of the production benchmark, split chronologically into calibration and held-out periods, we compare random sampling, historical caching, fixed representative subsets, and IRT-based adaptive testing. The results show that multidimensional 2PL adaptive testing achieves the best overall score fidelity: executing 200 questions, 38.5% of a full run, yields 1.03 pp of MAE. We nevertheless deployed difficulty-stratified fixed subsets because of their operational simplicity, and show they transfer without recalibration to five other agent families and remain stable across calibration windows as short as one day. Drawing on this deployment experience, we report practical recommendations for recurring production-agent evaluation.

Vision-language-action (VLA) policies based on flow matching generate action chunks through repeated evaluations of an action expert. Increasing the number of integration steps raises inference cost, but does not necessarily improve closed-loop success. We propose Coda, which reallocates part of this integration budget to a single learned endpoint correction. A frozen policy first completes a few-step noise-to-action trajectory; a lightweight Transformer then predicts a demonstration-supervised residual using the candidate action, source noise, and shared observation-prefix cache. Only the corrector is trained. On 50 RoboTwin Easy tasks, five-step Coda improves success from 71.64% to 74.68% over the matched five-step baseline, while reducing forward latency by 30.2% relative to the default ten-step policy. A two-step configuration achieves 71.88% success with a 2.12$\times$ speedup. An independent 13-task control shows a 5.69-percentage-point gain at nearly equal latency, supporting correction as an effective alternative to additional integration. The same design also improves frozen official SmolVLA, raising two-step success from 60.8% to 69.4%. These results show that endpoint correction improves the quality-latency trade-off of frozen flow-matching policies.

Generative world models offer controllable and repeatable closed-loop simulation for end-to-end and vision-language-action driving policies, but production deployment exposes three unresolved requirements: faithfully reproducing a mixed fisheye-pinhole rig at native resolutions; reconciling causal, per-timestep interaction with long-horizon stability and low latency; and preserving scene identity when a location is revisited. We present ZYT-World, a single architecture that natively generates four fisheye views with field of view > 180° and three pinhole views. Projection-specific Plucker adapters encode camera geometry, ego-motion adaptive layer normalization provides global motion control, and a lightweight pixel-aligned layout conditions traffic participants and signals through instance-level boxes, headings and colors. Heterogeneous training combines full-rig geometric coverage with high-resolution detail. Teacher forcing, causal consistency distillation, self-rollout distribution matching distillation, and RigCritic transform a 40-step bidirectional teacher into a one-step, per-latent streaming generator, with RigCritic evaluating the seven-view rig jointly. A 19M-parameter variational autoencoder decoder (TinyVAE), W8A8 quantization, and our inference engine reduce decoding, backbone, and incremental-execution costs, respectively. Finally, cross-trajectory pairs derived from real captures train a plug-in implicit-memory module that preserves place-specific evidence. On the internal multi-view test set, the one-step model retains more than 90% of the teacher's PSNR and SSIM, while FID, FVD, and LPIPS stay within 11% of the teacher. Under the generator-only timing in Figure 2, it is 107.7 times faster than the 40-step bidirectional teacher. TinyVAE decodes 59.8 times faster than Wan. 30s rollouts and cross-trajectory revisits show the intended long-horizon and memory behavior.

Quantization offers a practical path to deploying diffusion models with reduced memory and computation, but aggressive compression can cause quantized outputs to deviate substantially from their full-precision counterparts. Sampling-stage correction methods seek to compensate for such deviations during sampling, but existing approaches rely primarily on local information and underexploit trajectory history, limiting their ability to correct errors that propagate across timesteps. In this work, we formulate sampling with a quantized denoiser as an online estimation problem, using the history of quantized denoiser outputs to recover the underlying full-precision outputs required by the sampler. We propose QuAKE, a Quantization-Aware Kalman Estimator that combines a smooth trajectory prior with a conditional Gaussian observation model. At each sampling step, QuAKE recursively updates the posterior over the output window in closed form and feeds its posterior mean to the sampler. QuAKE is a lightweight plug-and-play corrector that requires no modification to the quantized network and naturally supports arbitrary high-order multistep ODE samplers. Experiments across W4A4-quantized text-to-image diffusion models show that QuAKE consistently outperforms existing methods in reducing the distributional discrepancy from full-precision sampling.

Speculative decoding accelerates LLM inference by drafting multiple tokens in parallel, with tree-based methods further improving efficiency through hierarchical structures. Dynamic-tree methods such as EAGLE-3 perform well under greedy decoding via deterministic top-K expansion and global pruning. However, in stochastic decoding (T>0), this mechanism collapses the draft distribution into one-hot probabilities, causing a severe drop in acceptance rate. This creates a dilemma: dynamic-tree methods sacrifice stochastic sampling to preserve context-aware topology, while static-tree methods preserve stochastic sampling with context-agnostic structures. The issue arises because the same probability distribution is used for two conflicting tasks: constructing the tree and verifying tokens. This coupling makes direct injection of randomness challenging due to the resulting stochastic process. We resolve this by decoupling these roles: RheoSampling assigns a token sampled from the draft distribution a proxy probability for tree expansion and pruning alongside its true sampling probability for verification. Specifically, we inject a sampled token among the deterministic top-K slots and treat it with different probabilities during construction and verification, making RheoSampling the first dynamic-tree method with both context-aware top-K construction and stochastic sampling while maintaining losslessness. We establish the lossless guarantee through an equivalence-class analysis that compresses the stochastic tree space into tractable classes. An OT-based verification strategy and a sparse draft mechanism ensure that theoretical gains translate into practical efficiency. Experiments across LLMs and benchmarks demonstrate improvements in acceptance rate and speedup over state-of-the-art dynamic tree methods. This framework may provide a template for analyzing stochastic tree structures.

Spatial NPUs such as AMD XDNA place compute tiles beside small local memories and leave data movement between them to software. Mapping a multi-stage workload onto such a device is largely a question of where the intermediate tensors live. We report what we learned making those choices for FlashAttention with the open-source IRON and MLIR-AIR flows. We compare four reference designs on XDNA 1 and XDNA 2: one runs each operator separately, two stream between operators on chip, and one fuses all three attention stages into a single kernel. The fused kernel holds the $\boldsymbol{QK}^{\mathsf T}$ scores in compute-tile local memory and reduces partial results over the cascade interconnect, so the scores never return to shared MemTile memory. On XDNA 2, it reaches 3.62 TFLOP/s over complete end-to-end execution, twice the IRON design, with 5.3 to 7.2 times the energy efficiency of the integrated GPU on the same chip at 2K tokens and above. It covers twelve LLM configurations, from BERT to DeepSeek, up to 128K tokens. Roofline analysis at each memory level explains this result and shows when to stop. XDNA 1 has lower ridge points, so streaming on chip already reaches the compute-bound regime: the same fusion that doubles throughput on XDNA 2 is nearly wasted on XDNA 1. Comparing a mapping's operational intensity against each level's ridge point predicts which case applies before writing any code. Fuse until the mapping clears that ridge point, then stop. We release the reference designs as maintained open source.

Large language models (LLMs) are moving onto mobile devices for increasingly diverse workloads over text, images, video, and audio. These applications often require long contexts, making the Key-Value (KV) cache a dominant memory bottleneck because it grows linearly with sequence length and is accessed at every decoding step. Prior work reduces KV-cache footprint through low-rank compression, token eviction, or flash offloading, but the resulting reconstruction overhead, irreversible token loss, or I/O stalls can offset the benefit of saving memory. We present TierKV, a mobile LLM inference framework built on Predictive Multi-Tier Cache Optimization (PMCO). Before decoding starts, PMCO predicts future cache demand from prefill hidden states and jointly assigns tokens to exact, low-rank, and flash-offloaded tiers under the device memory and accuracy budgets. This formulation retains access to the full context, removes the circular dependency of reactive eviction, and admits a closed-form solver that selects tier boundaries and per-layer ranks at runtime. Across eight text, vision, and audio models on three mobile SoCs, TierKV improves prefill throughput by up to 17.6x over existing mobile LLM frameworks, reduces RAM-resident KV cache by 12.5-34%, thereby enabling substantially longer contexts under the same memory budget, while incurring only minor accuracy degradation.

Running large language models (LLMs) locally continues to be limited by restrictions of compute and memory on consumer hardware. The popular acceleration technologies, such as quantization, speculative decoding, and adaptive inferencing, offer substantial speed boosts but usually necessitate retraining, per architecture tuning, or draft models. SpecQuant is a trainingfree framework, that combines speculative decoding with multiparent quantization to perform adaptive, efficient inference of LLMs. SpecQuant derives multiple quantized variants (INT4, FP8, FP16) from a shared base model, and dynamically routes queries based on predicted complexity; lightweight variants are used for simple or factual tasks, and full-precision models are used for complex reasoning tasks or long-context inputs. The shared-weight design of SpecQuant ensures sufficient token acceptance for speculative decoding without compatibility issues using separate draft parent models. We evaluate SpecQuant on Qwen2.5 based models on the MMLU, AlpacaEval, and GSM8K datasets, or benchmarks, demonstrating 35-43% speedups without degrading accuracy greater than 2%, substantial within the LLM community. SpecQuant enables practical on-device LLM deployment across diverse hardware without special infrastructure or expertise.

The growing elderly population gives rise to an urgent need for intelligent support systems, particularly for individuals with Mild Cognitive Impairment (MCI). This paper presents PV-Care, a proactive AI-driven assistance scheme that integrates wearable electroencephalogram (EEG) sensing with visual environmental perception to provide real-time, context-aware voice assistance for MCI users. Unlike traditional assistant systems that passively wait for user commands, PV-Care actively initiates helpful interactions based on the user's detected brain states, including Learning, Memory Recall, and Resting, using a novel deep neural architecture named Spatial and Frequency Refinement Network (SFR-Net). By combining EEG-based cognitive-state recognition with AI-based visual analysis, PV-Care generates structured "4W-UT" prompts to guide the output of large language models (LLMs). Simulation results and user studies validate the high accuracy of the proposed SFR-Net and the effectiveness of PV-Care's context-aware assistance. These results indicate that PV-Care is a feasible and promising solution for MCI caring.

The success of Large Audio Language Models has driven the development of massive multimodal networks exceeding billions of parameters. However, the demand for privacy-preserving, low-latency processing has shifted focus toward Small Audio Language Models (SALMs) capable of on-device execution. In this paper, we introduce Samsone, a family of SALMs designed for edge computing. Our core model, Samsone-134M, establishes a new state-of-the-art for its size class across multiple benchmarks. We further explore the scaling laws of SALMs by introducing Samsone-99M and Samsone-356M. Despite their compact footprint, the Samsone family delivers performance competitive with models orders of magnitude larger. To foster open research and reproducibility, we train Samsone on publicly available data. We release the training code, model weights, mobile-optimized checkpoints and provide an open-source Android application to demonstrate real-time on-device inference of Samsone.

Best-of-$n$ (BoN) sampling is a simple yet effective inference-time alignment method, but hard maximization provides only coarse control over the trade-off between reward and distribution shift. Soft Best-of-$n$ (Verdun et al. 2025) provides smoother control and converges to the optimal distribution associated with KL-regularized reward maximization. In this paper, we introduce ExpBoN, an alternative soft BoN method based on the exponential-noise report-noisy-max mechanism. It admits an exact finite-$n$ decomposition, which yields exponentially fast convergence in total variation, expected reward, and both directions of KL divergence. We provide comprehensive theoretical analyses of its convergence and regret behavior. We further integrate ExpBoN into the guided speculative inference (GSI) framework (Geuter, Mroueh, and AlvarezMelis 2025), resulting in ExpGSI, for efficient reward-guided LLM alignment. ExpGSI yields substantial reductions in computational cost while maintaining comparable accuracy. Experiments on MATH500, MMLU-STEM, and Minerva Math with the Qwen2.5-Math and Qwen3 model families show that ExpGSI reduces estimated computation by $14%$-$39%$ across candidate budgets for Qwen2.5-Math and by up to $45%$ at $n=16$ for Qwen3. Overall, our results provide a theoretical and algorithmic foundation for exponential-noise BoN and efficient test-time LLM alignment.

Large language models can hold knowledge they do not report. A model may sandbag on a capability evaluation, or answer against what it internally knows, and its outputs alone cannot tell whether it is hiding an answer or simply does not have one. We borrow the Concealed Information Test, a forensic method that identifies guilty knowledge by presenting a suspect with the true detail among plausible decoys and measuring a stronger response to the item they recognize. Our method, Probe of Internal Recognition (PIR), does the same inside a model. It presents a question with its candidate answers and reads, from the model's internal states, which candidate the model recognizes as correct. PIR is reference-free, needing no honest reference model and no labeled truth corpus. Across eight models from five families (Gemma, Qwen, Llama, Mistral, and Phi), PIR recovers the recognized answer at 0.70 to 0.87 balanced accuracy, well above the 0.28 to 0.40 unknown-item baseline and the 0.25 chance rate. It stays readable across every form of concealment we test, from prompted deception and trained sandbagging to external password-locked and circuit-broken checkpoints, with recognition between 0.85 and 0.93. When the model hides a known answer, recognition stays high. When unlearning removes the knowledge, recognition drops to the level of a question the model never knew. PIR therefore separates a model that will not answer from one that cannot, which supports sandbagging audits and unlearning verification. The signal is causal, adds information beyond black-box behavioral cues, and extends from multiple-choice questions to free-form generation.

Air operations rely on complex rules, established procedures, and time-critical analysis under limited connectivity and strict security constraints. In such environments, analysts must combine written doctrine with images, often without access to external computing resources. This paper studies offline large language models as decision support tools, deployed in isolated and restricted environments to give analysts access to doctrinal knowledge that remains traceable to its original sources through natural language interaction. We describe a modular retrieval-augmented architecture suitable for operation without Internet connectivity, supporting both text and image input from technical manuals. As a first step toward evaluating this architecture, we report a pilot study with four image analysts of the Brazilian Air Force, combining (i) a doctrinal knowledge assessment based on their electronic-target identification doctrine, comparing human and proposed system performance on the same test, and (ii) a measurement of the cognitive workload involved in manually producing a reconnaissance target report (Relatório de Missão de Reconhecimento - REMIR) without AI assistance. The results show a demanding manual task, especially in terms of mental demand (6.0/7) and effort (5.0/7), while the proposed system matches the human score (8/10) and completes the assessment in 7.1 minutes (compared to a human average of 26.5 minutes), establishing a baseline for future AI-assisted evaluation. Finally, we describe a future evaluation protocol to systematically compare manual and AI-assisted workflows.

Post-training weight-activation quantization reduces the memory and inference costs of large language models, but aggressive W4A4 quantization remains difficult because activation outliers degrade effective quantization resolution. Although weight optimization, channel-wise scaling, and orthogonal rotation mitigate this problem, the error components they address and their relationship remain unclear. Using an exact decomposition of local weight-activation quantization error into an activation-guided weight compensation term and an orthogonal residual, we bound the residual using persistent channel-wise outlier and regular activation quantities. This decomposition clarifies which error components can be addressed by weight compensation and which require transformation design. We then use the residual bounds to derive practical guidelines for applying randomized Hadamard rotation, sign selection, and channel scaling. In particular, the analysis explains how random signs suppress constructive interference among persistent outlier channels, how sampling multiple sign patterns can improve transformation selection, and how second-moment balancing leads to an $L_2$ scaling rule while a further relaxation recovers SmoothQuant-style $L_\infty$ scaling. We evaluate these guidelines through backpropagation-free configurations across eight Llama and Mistral models, obtaining performance competitive with gradient-trained SpinQuant.

Understanding and modeling human intelligence are parallel goals shared by artificial intelligence (AI) and cognitive science. As AI systems grow increasingly capable, in what ways do model responses resemble human responses, and where do they systematically diverge? The sheer breadth and diversity of the tasks humans can perform and think about pose a challenge for scalable and rigorous comparison between humans and models. We introduce CogGym, a scalable, unified framework grounded in cognitive science for systematically comparing model and human behavior on matched experimental trials. CogGym uses a semi-automated, human-in-the-loop pipeline to standardize diverse experimental paradigms into a task-agnostic Experiment Markup Language (EML), enabling reproducible and faithful comparison at scale. For initial release, we curate and standardize 258 cognitive experiments from 100 papers that focuses on human commonsense reasoning, and evaluate 50 large language models against human responses. We find a clear scaling trend where larger and more recent AI models better reproduce human judgments. Yet AI models' improvement on such common reasoning tasks is considerably slower than the gains observed on formal-reasoning benchmarks like math and coding, and model--human fit remains well below human splithalf reliability ($R^2 = 0.93$ on text, $0.95$ on image, and $0.92$ on video) with the best models achieving $R^2 = 0.59$ on text, $0.58$ on image, and $0.43$ on video experiments. We intend for CogGym to provide a living evaluation framework that continually incorporates new cognitive science experiments to characterize where model behavior resembles human behavior, where it systematically diverges, and how those patterns change as models and experiments evolve.

Activation steering has become a widely used approach for controlling language models during explicit chain-of-thought (CoT) reasoning, motivating its extension to latent CoT. However, we find that steering continuous thoughts produces substantially weaker effects on subsequent language generation than steering explicit CoT, even when the hidden representations are moved by comparable amounts. We first show that task information remains identifiable in continuous thoughts. Hence, we hypothesize a latent-to-language transition gap, in which an intervention effect in latent space fails to transfer to language generation. Two further results support this hypothesis: the output distribution changes abruptly at the transition boundary, and task-related directions exert much weaker bidirectional control in latent CoT than in explicit CoT. These findings identify the transition interface as a central target for evaluating and designing future latent-steering methods.

Homomorphic encryption (HE) has emerged as a promising approach to privacy-preserving machine learning (PPML), enabling computation directly over encrypted data. In HE-based PPML, a client submits an encrypted input to the server, which evaluates models such as large language models (LLMs) without access to the underlying plaintext. However, we identify a critical security vulnerability in this setting: HE-LLM inference is vulnerable to malicious clients that submit adversarial prompts, such as jailbreak attacks. The same confidentiality that protects benign clients also prevents the server from inspecting incoming prompts or generated responses, making adversarial attempts difficult to detect or block and potentially allowing successful attacks to remain entirely invisible to the server. To address this vulnerability, we propose HE-Guardrail, a framework that evaluates guardrail mechanisms entirely over encrypted data and homomorphically controls whether the target-model response is returned to the client. We instantiate HE-Guardrail with three representative guardrails - Llama Guard, JBShield, and GradSafe. Our results show that HE-Guardrail closely reproduces the decisions of the corresponding plaintext guardrails in the encrypted domain, with distinct security-efficiency-utility trade-offs.

Detecting pretraining data in large language models is challenging because high likelihood can reflect either training exposure or strong generalization. In the joint space of prediction loss and predictive entropy, a likelihood-only detector uses a horizontal boundary and can mistake predictable non-members for members. Motivated by this, we introduce an inclined boundary that evaluates prediction loss relative to predictive entropy. Our analysis shows that entropy correction can preserve the expected membership signal while reducing its variance, thereby improving standardized member--non-member separation. We further extend the mean--variance analysis to the more general setting with a nonzero mean entropy gap. Interestingly, this entropy-adjusted score admits a Helmholtz free-energy interpretation, leading to Energy Transfer Detection (ETD), which views pretraining data detection from a macroscopic residual free-energy transfer perspective. Extensive experiments show that ETD achieves the best average detection performance, improving average AUROC by up to 3.5% and TPR@5%FPR by up to 5.1%, while remaining robust across diverse settings.

Chain-of-Thought (CoT) reasoning has been shown to improve the performance of large language models (LLMs), yet existing optimization methods largely rely on outcome-based feedback, leaving the logical validity of intermediate reasoning steps largely unverified. To address the gap whereby LLMs arrive at correct final answers through logically flawed intermediate reasoning chains, we propose LogicTrack, a neuro-symbolic framework that audits reasoning trajectories by auto-formalizing each reasoning step into symbolic representations and verifying it with automated theorem provers. LogicTrack introduces Solver-Based Backtracking Reward (SBR), a step-wise scoring mechanism that quantifies logical soundness and guides backtracking tree search at inference time. We further extend LogicTrack to construct supervised fine-tuning (SFT) data with backtracking traces from its trajectories, enabling fine-tuned models to internalize step-wise auditing as an intrinsic capability. Extensive experiments across 8 reasoning benchmarks and 7 LLMs demonstrate that LogicTrack effectively improves both the verifiability of reasoning chains and final answer pass rate, thereby enhancing overall CoT quality and trustworthiness in high-stakes domains.

High-consequence subsurface decisions often rely on sparse data that permit competing geological interpretations. Determining consistency of these interpretations with the available observations remains challenging. We present a workflow that addresses this challenge by translating competing geologic interpretations into alternative priors and ranking them according to their consistency with hydraulic-head observations. A key step in this workflow is exploiting the broad knowledge of image-generation foundation models to transform nuanced geologic interpretations into data ready for computer modeling. For each interpretation, a text-to-image foundation model generates an ensemble of geologic images, and a separately trained variational autoencoder learns an interpretation-specific latent representation. A supervised inverse network maps head observations into this latent space, and the frozen decoder reconstructs an image that is mapped to a log-conductivity field. Steady-state flow simulations predict heads, and the aggregate normalized head error determines the ranking. We evaluate the framework using a synthetic benchmark based on the Johansen Formation with three interpretations of decreasing consistency with the reference geology. Across 595 test cases, the Precise & Accurate interpretation produces lower normalized errors than Accurate in 58.5% of cases and Mismatched in 82.5% of cases. Accurate outperforms Mismatched in 65.5% of cases. We then compare spatial representations of two published conceptual models of the Culebra Dolomite Member at the Waste Isolation Pilot Plant. The revised representation yields an aggregate normalized error of 7.598, compared with 8.595 for the original, consistent with the documented conceptual-model revision. The framework enables quantitative comparison of competing geological interpretations using available hydraulic observations.

We propose a Verification Reward Model (VRM) framework for training language models to create and repair chip verification artifacts. The central object is a versioned verification contract that binds requirements, permissible stimulus, observation boundaries, reference behavior, evaluation budgets, and acceptance criteria. Compiler, simulator, formal, mutation, coverage, and expert-review evidence are converted into auditable records. Deterministic acceptance checks remain outside the learned model. A learned outcome model predicts expensive future evidence from the specification, the generated artifact, and explicitly masked partial evidence; a semantic critic identifies evidence-supported weaknesses; a deterministic reward composer translates the resulting quality vector into task-conditioned training rewards. Extensions include paired clean/fault interventions, marginal fault-discovery rewards, uncertainty-aware evaluation scheduling, and a quarantined cross-domain learning loop. Evaluation emphasizes independently validated fault detection, false alarms, generalization across design families, and total cost to reach a specified quality level. We describe a first experiment on a small, open-tool-compatible benchmark with lightweight models, with full UVM capability admitted through feature-specific qualification. This is a position paper: we specify the framework and the experiments that would test it, and we report no training, EDA, or silicon results.

Large Language Models (LLMs) are increasingly deployed in applications that must weigh clashing moral values, yet even strong models exhibit hidden biases and brittle instruction-following across languages. We introduce a 12,000-instance dataset of two-option dilemmas covering pairwise three value conflicts: Honesty vs. Justice, Justice vs. Autonomy, and Autonomy vs. Honesty, along with their translations into Hindi, Arabic, Spanish, and Chinese, to probe cross-lingual behavior. Benchmarking on GPT-5-mini reveals that it consistently favors Honesty over Autonomy across all five languages when no policy is given. The Llama-3.2-1/3B models exhibit strong first-option bias; however, both plain fine-tuning and Direct Preference Optimization fine-tuning effectively remove this bias, increasing accuracy to greater than 98%. In order to decouple the effect of learning correlations in the dataset from abstract values, we propose a task vector transfer based experiment where after computing the task vectors for a direction of value preference we orthogonalize it with respect to the general instruction following vector. Our experiment shows that this method is effective in isolating the direction of the specific value preference that can successfully be used to conduct task arithmetic to obtain a model with the opposite stance.

Reinforcement learning (RL) controllers have been recently adopted for Unmanned Aerial Vehicles (UAV) navigation and control. However, they are susceptible to action-space attacks that overwrite the action commands after the policy generates them and before the actuators execute them. While most existing defenses target attacks on the policy's inputs, those addressing action-space attacks retrain the policy at training time and are not resilient to corrupted actions at runtime. We propose ASGARD, a two-phase teacher-student pipeline for making RL-based UAV control resilient to action-space attacks. In the teacher phase, an encoder combines the UAV's physical state with action-attack-related privileged information to produce an action-attack-aware latent that trains the RL control policy and a monitor that outputs corrected action commands to the actuators. In the student phase, both the encoder and the monitor are trained via supervised learning from their teacher counterparts to run on-board using only the UAV's physical state history. We evaluate ASGARD across attack scenarios targeting different action commands on UAV. We find that ASGARD is resilient to action-space attacks and completes the missions despite the attack. We further find that ASGARD generalizes to unseen attacks and remains resilient against stealthy attacks.

During manipulation, robot and scene motion can move previously observed regions outside the camera's field of view. Geometry-aware RGB features encode visible structure, while control under partial observability requires scene memory that integrates observation history and grounds inferred content in current evidence. We introduce \lifd{} (Look, Imagine, Focus, and Do), a framework for persistent, 3D-aware scene memory. LIFD learns scene tokens through multi-view agreement, then completes them from a single RGB view and recurrent memory using rectified flow. Anchor-Guided Cross-Attention anchors generation to current geometry-aware features, and compact slot features condition a visuomotor policy. Multi-view and geometric supervision are used during representation learning; deployment requires one RGB camera, proprioception, and a task instruction. LIFD (Staged) reaches 91.6% average success on LIBERO and 79.8% on MetaWorld, improving LIBERO average success by 11.1 percentage points over Joint training. After policy-head adaptation with ten demonstrations per family, LIFD achieves 56.0% mean success across four UR5e task families, compared with 40.5% for OpenVLA-7B.

Vision-language-action (VLA) models have demonstrated strong capabilities in robotic manipulation, yet their reliance on visual perception limits robustness in contact-rich environments, where critical physical interaction states may not be visually observable. Existing tactile-enhanced VLA methods improve physical grounding using observed tactile feedback, but most remain largely reactive rather than explicitly modeling how contact may evolve. Therefore, we propose ForeTac-VLA, a forecasting-based tactile-vision-language fusion model that predicts future tactile states to guide action generation. Specifically, ForeTac-VLA encodes recent tactile observations into temporal representations and integrates them with vision-language features through bidirectional cross-attention. Further, a transformer-based forecasting module predicts multi-step future tactile states, enabling the model to reason jointly over observed and anticipated contact. Finally, the fused multimodal representations and predicted future tactile states are fed into the VLA backbone to condition action generation. To stabilize training, a ground-truth-to-prediction curriculum is employed when early forecasts are unreliable. Across four real-world contact-rich manipulation tasks, ForeTac-VLA achieves an average success rate of 95%, outperforming the fine-tuned VLA model by 36.25 percentage points and state-of-the-art tactile-enhanced VLA baselines by over 22 percentage points. ForeTac-VLA also maintains strong performance under low-illumination and visually cluttered conditions. Video demonstrations can be found on https://foretac-vla.github.io/