Cryptography and Security
ALIBI: Adversarial Legitimacy Injection in Binary Input against LLM Malware Analyzers
Hyeongjun Choi, Wonyoung Jung, Haehoon Seo et al. (+1 more)
Large
language models are being integrated into malware triage workflows as
reasoning components that summarize static evidence and produce analyst-facing verdicts. This paper shows that the same
reasoning capability introduces a new attack surface. We present ALIBI, a semantic cover story attack against frontier LLM-based malware analyzers. ALIBI adds a small, non-executed read-only section to a compiled binary, containing a coherent but false security product narrative, without altering imports or executable behavior. Instead of issuing direct instructions to the model, it reframes suspicious evidence as expected behavior of a benign endpoint security tool. On a frozen PE set of 50 malicious samples, the payload flips 30 of the 35 baseline-malicious samples to benign on Gemini 2.5 Pro, while GPT-5.5 Pro and Claude Opus 4.7 produce substantial severity downgrades with significant confidence reductions even when verdict labels are preserved. The attack transfers to ELF binaries, where Gemini flips 16 of 40. A verification-guided defense
prompt roughly halves the benign verdicts, but 42.9 percent of malicious samples still reach benign. LLM malware analyzers therefore require provenance checks that separate verified facts from attacker-controlled claims, not narrative trust.
Read the full abstract ▾
All authors Hyeongjun Choi, Wonyoung Jung, Haehoon Seo, Sungyup Nam
Artificial Intelligence
LearnActCoder: Role-Aware Error Memory for Adaptive Clinical Coding Agents
Meysam Ghaffari, Bhaskar Sen, Nasim Sabetpour et al. (+3 more)
Clinical coding
agents repeatedly encounter the same failure modes, including unsupported codes, missed documented conditions, specificity errors, and procedure-coding convention mismatches. We introduce Learn-Then-Act, an inference-time adaptation framework that converts errors from a small labeled LEARN batch into a structured Mistake Knowledge Database (MistakeKDB). False-negative lessons are routed to a recall-oriented Coder, while false-positive lessons are routed to a precision-oriented Judge. We instantiate the framework in LearnActCoder, a Coder-Judge clinical coding pipeline with lookup-table grounding where available. On 150 matched MIMIC-III notes, structured MistakeKDB improves CPT F1 by 5.9 percentage points, while raw-example and reflection-style
memories remain near the no-
memory baseline; the ICD-9 improvement is not significant. On a matched MIMIC-IV cohort,
memory shifts ICD-10 coding toward higher precision at a recall cost, leaving F1 statistically unchanged. Applying the same
memory to 1,000 held-out MIMIC-III notes maintains a stable ICD operating point, providing scale/stability evidence. Overall, the results are consistent with structured, feedback-derived error
memory being useful for adapting clinical coding behavior across cases without weight updates or changes to the underlying workflow. Absolute CPT/HCPCS performance remains low, and the system is evaluated retrospectively rather than in clinical deployment.
Read the full abstract ▾
All authors Meysam Ghaffari, Bhaskar Sen, Nasim Sabetpour, Nina Fatehi, Animesh Agarwal, Carlos Morato
Cryptography and Security
Reachability, Not Observation: Containing Systems Whose Wiring Changes
Yoshiaki Takashita
Containment decisions – where to put a firewall, which links to monitor, what a program may reach – are computed from an observed structure, and observation is a snapshot. We ask what a snapshot misses when the wiring changes over time. On a hypercube whose active dimension rotates, a balanced split shows zero crossing edges at 93% of instants, yet 8,192 edges must be blocked permanently; adding one always-on ring, a defender sees 2 where 8,194 must be blocked, a factor of 4,097. A time-aware defender holds 585 blocks on average, but one step of clock lag drops its containment to 0%. On the real Internet the same gap is only x1.8-2.0 (1997-2000) and x1.5-1.6 (2024-2026) once growth is removed, so the blind spot is introduced by design, not inherited from the world – and it has been designed: the round-robin schedules of optical datacentre fabrics have a gap equal to their period. A declared capability map, checked by static reachability over a real application's call graph, catches all 8 planted holes; the string deny-list previously in place catches 2. One calculation generates every number from one parameter, the period, and reads three boundaries not usually called schedules: frequency hopping, whose standard results are these closed forms with channels in place of edges; the air gap, whose always-on crossing set is empty and whose known breach came at the one phase a snapshot misses; and the tool surface of a coding
agent, inventoried from the inside. Turning the cuts on that
agent: idle, its instantaneous state cut is zero, while 6 channels carry it across a context reset, none of them the network, so severing the network removes 0. The channel that spawns copies is a branching process with a sharp threshold at approval rate 1/b, below which denial is unnecessary and above which denial is insufficient. Contain by the paths that exist, not by the behaviour that was seen.
Read the full abstract ▾
Computer Vision
SeetaPsych v1.0: An Open-source Computer Vision Toolkit for Behavior-based Psychological Measurement
Jiabei Zeng, Chiqin Li, Kaizhou Li et al. (+7 more)
Automated visual analysis opens new avenues for behavior–based psychological measurement. Nevertheless, existing technological modules are typically scattered across task specific systems with heterogeneous interfaces and disparate deployment requirements. In this work, we present SeetaPsych v1.0, an open source, unified and extensible computer
vision toolkit designed to extract psychologically relevant signals from facial images and/or face based
videos. The current release encompasses four major core modules aiming at behavior–based physiological perception: unified face based
emotion analysis (simultaneous facial expression recognition, facial action unit detection, and valence–arousal estimation), camera based heart rate estimation, screen point–of–gaze estimation, and scene gaze following. A suite of auxiliary preprocessing modules for human centric visual analysis is also included, comprising face detection, facial landmark detection, and head detection. These functionalities are encapsulated within a modular Pipeline/Runner architecture that automatically resolves attribute dependencies, constructs computation graphs, and support intermediate result sharing among modules. SeetaPsych provides standardized Python APIs to facilitate reproducible, large scale analyses, alongside an interactive WebUI for rapid, code–free method evaluation. Overall, SeetaPsych offers an integrated and accessible visual measurement platform for research in psychology, behavioral science, human computer interaction, and related fields.
Read the full abstract ▾
All authors Jiabei Zeng, Chiqin Li, Kaizhou Li, Fei Chang, Yong Li, Yuanhao Zhao, Dan Han, Wenqiang Yang, Xilin Chen, Shiguang Shan
Machine Learning
Learn Your Own Thoughts: Abstract Token Curriculum
Khashayar Gatmiry, Avrajit Ghosh, Parsa Mirtaheri et al. (+4 more)
Large
Language Models (LLMs) have achieved remarkable
reasoning capabilities by utilizing chain-of-thought (CoT) as a scratchpad for intermediate stages of thinking. However, CoT techniques require explicit supervision on thinking
tokens, which requires rich, task-specific data. In this work, we propose Abstract
Token Curriculum (ATC), a novel curriculum learning framework that elicits effective continuous intermediate representations without direct supervision or manual scratchpad design. ATC gradually increases problem complexity through a sequence of distributions, training the model to develop internal abstract “thoughts” in the continuous representation space. This paper provides both theoretical and experimental evidence for the benefits of ATC and its advantages over previous methods for training continuous thoughts. Theoretically, we show that for learning parity functions with single-layer softmax
attention using ATC,
attention naturally focuses on the CoT
tokens in the context that provide the “easiest path” to predicting the next
token. Experimentally, we show ATC's effectiveness on graph reachability and arithmetic learning tasks.
Read the full abstract ▾
All authors Khashayar Gatmiry, Avrajit Ghosh, Parsa Mirtaheri, Jason D. Lee, Nika Haghtalab, Emmanuel Abbe, Peter Bartlett
Computer Vision
GAPrompt++: Multi-Granular Geometry-Aware Point Cloud Prompt for 3D Vision Model
Zixiang Ai, Zhenyu Cui, Yufei Guo et al. (+4 more)
Pre-trained 3D
vision models have substantially advanced point cloud analysis, yet adapting them to downstream tasks via full fine-tuning is computationally expensive and storage-intensive. Parameter-Efficient Fine-Tuning (PEFT) offers a promising alternative by reducing both adaptation cost and storage burden. However, existing prompting-based approaches ignore the intrinsic
geometric structures of point clouds, thereby limiting their adaptation capability. This limitation stems from their inability to encode both fine-grained
geometric cues and coarse-grained structural semantics, as well as failing to propagate such information effectively through the model hierarchy. To address these challenges, we propose GAPrompt++, a multi-granular geometry-aware prompting method that provides richer
geometric guidance for efficient 3D task adaptation. Specifically, we introduce a Point Shift Prompter that extracts multi-granular
geometric features across different scales, enabling instance-specific
geometric adjustments during adaptation. Next, a Keypoint Prompter adaptively generates point-level
prompts to highlight local
geometric saliency and fine-grained structural details. Furthermore, a
Prompt Propagation mechanism injects these multi-granular
geometric cues throughout the feature extraction hierarchy, strengthening the ability to capture essential
geometric characteristics. Extensive experiments show that GAPrompt++ achieves state-of-the-art performance among prompting-based PEFT methods and even surpasses full fine-tuning across diverse
benchmarks, while requiring less than 2% trainable parameters. In addition, to address the saturation of existing evaluation
datasets, we construct two more challenging
benchmarks derived from 3D Gaussian Splatting and Multi-View Stereo reconstruction, offering diverse and realistic point cloud scenarios to promote future research.
Read the full abstract ▾
All authors Zixiang Ai, Zhenyu Cui, Yufei Guo, Wenwen Qiang, Lei Chen, Jiwen Lu, Jiahuan Zhou
Data Structures and Algorithms
Improved Algorithms for Beck–Fiala with Bounded Sets
Dylan J. Altschuler
We give an efficient
algorithm with improved algorithmic guarantees for the (offline) Beck–Fiala problem when the sets have bounded size. Let $A$ be an arbitrary
matrix $A\in\{0,1\}^{m\times n}$ with at most $d$ ones per column and at most $s$ ones per row. Let $\log^*$ denote the iterated logarithm and $\ell_j$ denote the $j$-fold composition of log. Assume $s\le\exp(O(\sqrt d))$. We provide an efficient
algorithm that, for arbitrary sparsity $d$, gives $O(\sqrt d(1+\log^*n))$ discrepancy. Moreover, if $d\ge\ell_j(n)$ for a fixed integer $j\ge1$, the
algorithm gives $O_j(\sqrt d)$ discrepancy. The proof is a bootstrapping scheme using the Bansal-Jiang
algorithm.
Read the full abstract ▾
Software Engineering
A Closed-Loop Control Architecture for Reliable Constraint Satisfaction in LLM Text Generation
Quan Zhou, Shahbaz Siddeeq, Mika Saari et al. (+1 more)
Software systems increasingly embed a large
language model in features that must satisfy a numeric output constraint, that is, a requirement expressible as a number or an interval and checkable by code, such as a target word count or a target readability grade band. Because such a model is non-deterministic, is configured through natural-
language instructions rather than a typed interface, and satisfies a stated requirement only approximately, a single
prompt neither reliably meets the target nor preserves the source content. This paper presents and evaluates a closed-loop control architecture for this problem. It has five stages: generate, evaluate, adjust, archive, and analyze. The model is called only to write and to edit text, while deterministic code compares a composite readability value against a target band, rejects any edit that drops source entities, numbers, or keywords, and makes every accept decision. Over 114 single-shot generation jobs and 240 closed-loop runs on four commercial models, single-shot prompting met the target in 21.1 to 31.6 percent of cases and the closed loop in 92.5 to 98.8 percent, within two edit rounds on average and at a recall-based fidelity of 0.92 to 0.93; the two models common to both settings show the same effect. Because the controller optimizes the value on which success is scored, the result establishes reproducible control over a declared, computable metric and not validated human difficulty. The transferable practice is to declare the acceptance condition as code, bound the model to local edits, and gate every edit on a content check.
Read the full abstract ▾
All authors Quan Zhou, Shahbaz Siddeeq, Mika Saari, Pekka Abrahamsson
Machine Learning
Odds-Ratio Thompson Sampling: A Specification and Design Guide for Contrast-Based Multi-Armed Bandits
Sulgi Kim
Batched multi-armed bandits update on a service's own schedule, and the usual implementation carries each arm's absolute reward rate from one update to the next. When the shared level moves between batches, that
memory goes stale even though the comparisons between arms may not have. Odds-Ratio Thompson Sampling (OR-TS) instead carries the joint posterior over log-odds contrasts and fits the common level afresh in every batch, marginalizing it out. This paper specifies that update, places it inside a
Bayesian bandit
agent with two controls, decay for how much past evidence survives an update and aggressiveness for how sharply belief becomes allocation, and evaluates it against absolute-rate
memory. Across 86 public A/B series the level varies about twenty-five times more than the contrast. In prespecified synthetic environments a moving level costs absolute-rate
memory five times the regret and leaves the best arm below a majority of traffic in 7 of 20 runs, against none for OR-TS. In a policy simulation built from 71 real experiments, where the contrasts are too small to resolve, expected-click differences stay within 0.1% for 58 of them, yet contrast
memory still ends on the better arm more than twice as often. Where the contrasts themselves move, the bet fails, and that case is reported too.
Read the full abstract ▾
Information Theory · Cross-listed from stat.ME
Federated multiple testing with quantized p-values under family-wise error control
Prasanjit Dubey and Xiaoming Huo
Several institutions test one family of hypotheses under family-wise error control but cannot pool their data, so each site releases only a quantized report of its p-value for each hypothesis, such as whether the p-value is significant. The report's resolution is the number of values it can take. We quantify the power lost to such reports relative to the most powerful centralized procedure given all p-values (the oracle), with reports and combining rule chosen optimally. A report from which the p-value can be recovered loses no power, so the loss is due to compression, not decentralization. Under the stated regularity and power-objective conditions, every finite resolution loses power, and the optimal loss decays as the inverse square of the resolution: equal-width intervals attain this order, and no partition into as many
cells improves it. In two-site Beta examples, optimized one-bit reports are nearly lossless for a single hypothesis but lose several times as much power for two. At level 0.05 with two Beta(1,2) hypotheses, optimally combined three-bit equal-width reports retain 99.2% of oracle power. We give the power-optimal combining rule for given reports, and a rule using only their exactly known null distribution, valid under arbitrary dependence across hypotheses.
Read the full abstract ▾
Game Theory · Cross-listed from math.OC
Optimal Strategies in a Sequential Contest with Failures
Alec Pannunzio, Ashley Shaffer, Gregory Shaver et al. (+1 more)
We study a Stackelberg model where two players compete in a sequential contest, and the player that successfully invests the most effort wins. Our formulation considers a probability that each player fails to invest their chosen level of effort, as captured by a failure function that increases with the amount of effort. The players each have their own cost of failure, as well as a maximum level of effort (capturing different abilities of the players). We prove three key results. First, we provide the optimal strategy for Player 1 (the follower), to maximize their utility given an observed effort exerted by Player 2 (the leader). Next, we characterize the optimal strategy to maximize the expected utility of Player 2, anticipating the optimal response by Player 1. Finally, we analyze the expected utilities of each player when both play their optimal strategy, and provide key insights into how heterogeneous player abilities and failure costs affect the outcomes of the game.
Read the full abstract ▾
All authors Alec Pannunzio, Ashley Shaffer, Gregory Shaver, Shreyas Sundaram
Cryptography and Security
Trading Agents or Market Crashers? Dissecting Robustness and Security Failures in Academic Financial LLM Trading Schemes
Mengxiao Wang and Nitesh Saxena
Large
language model (LLM)
agents already trade real capital, and builders copy a few published designs that share models and news feeds, so one shock reaches many
agents at once. We present FARSIGHT (Financial
Agent Robustness and Security Investigation and Global Holistic Taxonomy), a systematization that rates trading
agents on robustness to
market stress and on security against source manipulation,
agent compromise, and
agents acting as attackers, and checks the ratings against released code. Across 15 academic schemes, no paper describes a halt that skips LLM inference or combines source provenance with cross-source corroboration, and the 7 code releases enforce no stop-loss or position limit outside the LLM. Three LLMs sell whole positions on a fake headline and on real hoaxes, including the fake press release behind the 2000 Emulex crash. In a simulated
market driven by these decisions, one fake headline crashes the
market in all runs with 400
agents, and 400
agents that sell when prices fall nearly double a 2010-sized crash. The pause that U.S. exchanges apply to large stocks starts in no headline run, whereas automatic halts and hard loss limits outside the model remove both effects. Restated as a domain-independent threat model, FARSIGHT maps the same gaps onto
agents in healthcare, law, software, and crypto, where compromised
agents spread harm as worms and as cascades. We derive 14 open problems and urge deployers to corroborate sources, keep interrupts that bypass the model, and enforce loss limits outside it.
Read the full abstract ▾
Discrete Mathematics · Cross-listed from math.CO
On integral polytopes related to Edmonds' problem
Hiroshi Hirai
In this paper, we study polyhedral aspects on commutative and noncommutative Edmonds' problems for computing the rank of linear symbolic
matrix $A = \sum_{k=1}^m A_k x_k$. We regard them as linear optimization over integral polytopes ${\cal P}(A)$ and ${\cal Q}(A)$, respectively, which are obtained by the convex hulls of exponent vectors of subdeterminants of~$A$ and its blow-ups $A^{\{d\}} = \sum_{k=1}^m A_k \otimes X_k$ $(d=1,2,\ldots)$. By extending previously known results on nc-rank, we establish a hierarchy of integral polytopes ${\cal P}(A) \subseteq {\cal P}^{\leq 2}(A) \subseteq {\cal P}^{\leq 3}(A) \subseteq \cdots = {\cal Q}(A)$ and show that the integrality gap of ${\cal Q}(A)$ relative to ${\cal P}(A)$ is at least $1/2$. Further, we show that if each $A_k$ is rank-2 skew-symmetric, then the above hierarchy terminates at the second level and the integrality gap is improved to $2/3$.
Read the full abstract ▾
Computer Vision
Understanding and Exploiting Diagonal Attention Sparsity in Autoregressive Image Generation
Daeun Kim, Junwha Hong, Changhun Oh et al. (+3 more)
Autoregressive image generation has emerged as a paradigm for
multimodal AI systems due to its compatibility with
transformer-based LLM serving infrastructures. However, generating thousands of visual
tokens per request makes decoding increasingly bottlenecked by KV cache accesses during
attention computation. Sparse
attention is particularly attractive for this workload because many visual generation applications tolerate moderate quality degradation in exchange for improved performance and efficiency. While sparse
attention has been extensively explored for text-based LLM inference, it remains unclear whether its sparsity assumptions generalize effectively to autoregressive image generation. We present the first systematic characterization of
attention sparsity in autoregressive image generation across diverse workloads and representative open-source models. Our analysis reveals several distinguishing properties, including a pronounced prefill-decode asymmetry, strong
attention concentration on
prompt and local
tokens, and a unique diagonal
attention sparsity pattern arising from the spatial locality of visual
tokens. Motivated by these observations, we propose a diagonal-aware sparse
attention mechanism that selectively skips KV entries along the diagonal
attention direction within a recent window. Implemented on top of a GPU-based serving system using FlexGen, FlashAttention-2, and custom kernels, our approach achieves up to 3.1x throughput and 1.19x latency improvements with less than 2% quality degradation compared to dense inference.
Read the full abstract ▾
All authors Daeun Kim, Junwha Hong, Changhun Oh, Yoonsung Kim, Yoonhyeong Lee, Jongse Park
Software Engineering
BA-TRACE: Boundary-Aware Trace Reconstruction for Scenario-Based Evaluation of Mixed AUTOSAR Adaptive and ROS 2 Vehicular Embedded Systems
Shunsuke Ito, Ryudai Iwakami, Hiroyuki Hanyu et al. (+2 more)
Modern vehicular embedded systems increasingly combine ROS 2-based autonomous-driving stacks with AUTOSAR Adaptive Platform (AUTOSAR AP). Such mixed stacks make scenario-based evaluation hard to interpret because execution paths cross DDS-SOME/IP middleware boundaries between ROS 2 and AUTOSAR AP. Existing simulators and tracing tools execute scenarios or collect platform-local traces but cannot reconstruct cross-domain data flows. This paper presents BA-TRACE, a boundary-aware trace reconstruction framework for scenario-based evaluation of mixed AUTOSAR AP and ROS 2 vehicular embedded systems. BA-TRACE combines ROS 2 trace events, AUTOSAR ara::log events, ARXML-derived structural dependencies, and bridge-level instrumentation to reconstruct an end-to-end execution graph across the DDS-SOME/IP boundary. A case study with an AWSIM/OpenSCENARIO-based object-detection and braking scenario shows that BA-TRACE reconstructs the expected cross-platform path and exposes boundary-specific latency such as point-cloud transfer overhead. The reconstructed
topology is used as evidence of traceability, not as proof of behavioral correctness or safety.
Read the full abstract ▾
All authors Shunsuke Ito, Ryudai Iwakami, Hiroyuki Hanyu, Tasuku Ishigooka, Takuya Azumi
Machine Learning
Opinion Dynamics-based Coalition Formation for Federated Learning in Heterogeneous IoT Systems
Mohammed El Hanjri, Anas Abouaomar, Hamidou Tembine et al. (+1 more)
Federated learning (FL) enables
privacy-preserving, on-device training across heterogeneous Internet-of-Things (IoT) deployments such as smart-city water-metering networks, where each smart meter observes a household-specific consumption time series. Under such statistical heterogeneity, the standard
Federated Averaging (FedAvg) aggregation averages dissimilar local models into a single global model that may fail to capture client-specific patterns. We address this by forming client coalitions directly in the local-weight space and aggregating at the coalition level. Extending a prior weight-driven coalition-formation scheme, we model coalition formation as a Hegselmann-Krause (HK) bounded-confidence opinion-dynamics process acting on the local weights, and develop variants of the HK interaction based on Euclidean-distance and cosine-similarity confidence criteria. The framework is applied to short-term water-consumption
forecasting with local Long Short-Term
Memory (LSTM) models and evaluated against FedAvg, Per-FedAvg, FedProx, and FedAvg with Euclidean-distance or cosine-similarity coalition formation. Experiments on a real smart-metering
dataset of water consumption show that the proposed HK-based coalition formation produces stable, endogenous coalition structures within at most ten inner iterations, incurs no additional client-side computation or communication compared to FedAvg, and reduces the average MAE by up to 54% relative to FedAvg, 39% relative to FedProx, and 24% relative to Per-FedAvg, while achieving the highest global accuracy (83-85%).
Read the full abstract ▾
All authors Mohammed El Hanjri, Anas Abouaomar, Hamidou Tembine, Abdellatif Kobbane
Computer Vision
IMFD: End-to-end Multi-Face Forgery Detection through Instruction-based Large Vision-Language Models
Dasom Choi, Sangjun Moon, Hyeongchan Im et al. (+5 more)
The rapid increase of deepfakes has raised significant concerns due to their spread on social media. Traditional multi-face forgery detectors crop and verify each face independently, ignoring background context and inter-face relationships, which often yields suboptimal performance. To overcome these limitations, we leverage instruction-based Large
Vision-
Language Models (LVLMs), which can interpret entire images and follow complex textual instructions. We propose a simple yet effective single-stage multi-face forgery detector, called IMFD (Instruction-based Multi-face Forgery Detector), which is trained end-to-end to jointly localize faces and predict per-face forgery labels. Rather than treating face box prediction only as a joint objective, IMFD explicitly integrates predicted face bounding boxes into the instruction as visual cues that enhance instruction grounding and forgery detection. To support the training and evaluation of IMFD, we convert existing multi-face forgery
datasets into an instruction-based format. Experimental results and analyses show that IMFD improves multi-face forgery detection by integrating face bounding boxes into the instruction, and consistently outperforms various state-of-the-art methods.
Read the full abstract ▾
All authors Dasom Choi, Sangjun Moon, Hyeongchan Im, Jaeeon Park, Jingun Kwon, Hidetaka Kamigaito, Taro Watanabe, Manabu Okumura
Robotics
UniExo: Unified Multi-Skill Policies for Musculoskeletal Locomotion and Co-Adaptive Exoskeleton Control
Yifei Yuan, Jakob Wolf, Ghaith Androwis et al. (+1 more)
Daily locomotion encompasses diverse activities and frequent transitions between them, yet most exoskeleton controllers are designed for a single activity or a narrow set of related movements. Changes in activity therefore typically require explicit mode switching and separately tuned or retrained controllers. Simulation-based learning reduces the need for hardware-based tuning but generally retains this limitation. Here we present UniExo, a framework that first constructs a multi-skill musculoskeletal human policy and then jointly trains an exoskeleton control policy with it. Four single-skill imitation experts for walking, turning, running and backward walking are distilled into a single network structured by a skill latent and subsequently fine-tuned through reinforcement learning on transition sequences. The resultant unified human policy achieves a mean tracking success rate of 94.7% on unseen clips of the four skills and exhibits greater robustness to perturbations than its constituent experts. A single hip exoskeleton controller (UniExo) is initialized from hip moment prediction of the human policy and co-adapted with it through multi-
agent reinforcement learning across the four skills. This co-adaptation shifts the timing of the assistance torque and raises the fraction of positive work delivered to the hip. When deployed on a custom hip exoskeleton, the controller generalizes across four treadmill speeds in six participants and assists one participant through a continuous route of all four skills and their transitions, without skill labels or explicit mode switching. UniExo thus provides a step towards replacing activity-specific controllers with unified, user-specific controllers that support diverse locomotor activities and the transitions between them.
Read the full abstract ▾
All authors Yifei Yuan, Jakob Wolf, Ghaith Androwis, Xianlian Zhou
Robotics
LYRIC: Language-Driven Physics-Based Character Control for Contact-Rich Whole-Body Object Interaction
Zeyu Han, Zichong Meng, Julian Tanke et al. (+8 more)
We present LYRIC, a generative flow-matching controller for
language-driven physics-based contact-rich interaction control, that enables simulated characters to perform contact-rich whole-body object interactions from a free-form
language instruction and a sparse terminal object goal. To obtain reliable expert trajectories from imperfect motion-capture references, a single tracking policy is trained using geometry-conditioned interaction rewards and relaxed reference tracking near hand-object contact. To guide interaction progress without prescribing a full-body kinematic reference, we factorize the controller into a task-level planner that predicts short-horizon object and
humanoid-root trajectories, and an action generator that resolves whole-body motion and contacts in closed loop. After behavior cloning, we freeze the planner and post-tune the action generator on policy using the planner's predictions as stable supervision for intermediate task progression. In a controlled OMOMO evaluation, our tracker achieves 64.3% success compared with 53.2% for an InterMimic reimplementation, while a unified policy achieves 76.5% on the full OMOMO
dataset. On the held-out split, LYRIC achieves 90.3% task success, compared with 74.2% for the strongest matched kinematic-planner baseline, with better semantic
alignment and motion quality. Without retraining, the controller also supports test-time object-waypoint guidance. Qualitative results further demonstrate robust, natural contact-rich interactions and zero-shot transfer to novel object shapes. The webpage is available at https://neu-vi.github.io/LYRIC/
Read the full abstract ▾
All authors Zeyu Han, Zichong Meng, Julian Tanke, Minami Matsumoto, Sergey Bashkirov, Yingruo Fan, Selim Engin, Dongseok Shim, Takashi Shibuya, Yuki Mitsufuji, Huaizu Jiang
Computational Complexity
Almost Optimal FPT Inapproximability for k-SetCover
Venkatesan Guruswami and Xuandi Ren
We show that $\bigl(\frac{\log n}{\log\log n}\bigr)$-approximate parameterized $k$-SetCover is W[1]-hard, and has no $n^{o(k/\log k)}$-time
algorithms under ETH. This improves upon the previous best factors $\bigl(\frac{\log n}{\log\log n}\bigr)^{1/k}$ in (Lin, 2019) and $(\log n)^{1/\operatorname{poly}(k)}$ in (Karthik, Laekhanukit, and Manurangsi, 2019). Here $k$ is the yes-case guarantee and $n$ is the number of candidate sets. While the best approximation ratio is still $O(\log n)$ via the greedy
algorithm, closing this $1/k$ gap in the exponent has been a longstanding open problem; we remove this loss via a simple direct reduction. The construction is self-contained and does not rely on the parameterized inapproximability hypothesis (PIH). Starting with sparse parameterized 2-CSP instances (Karthik, Marx, Pilipczuk, and Souza, 2024), we build a monotone CNF formula, which is equivalent to a SetCover instance. To obtain a $k$-versus-$h$ gap, the reduction enumerates all hash functions from $\Sigma$ to $[2h]$ and all unsatisfiable 2-CSP instances on the same constraint graph with alphabet $[2h]$. For each such instance, it asks for a certificate that the hashed label pairs are not all contained in that instance. Perfect hashing makes this enumeration efficient for $h=\log n/\log\log n$.
Read the full abstract ▾
Hardware Architecture
MiX: Micro-Inverted-Scaling for End-to-End Low-Bit Vision-Language Model Acceleration
Yuan Liao and Jae-sun Seo
The deployment of
Vision-
Language Models (VLMs) on edge devices is severely bottlenecked by
memory bandwidth, necessitating aggressive sub-8-bit quantization. Since edge accelerators are strictly constrained by area and power, they require end-to-end quantized models. However, the extreme dynamic range gap between multi-modal
tokens causes standard block formats to suffer "microscaling collapse," where a single massive outlier hijacks the shared exponent, underflowing surrounding elements and destroying
attention maps. To break this bottleneck, we propose Micro-Inverted-Scaling (MiX), a novel format that mathematically inverts the microscaling paradigm: rather than grouping multiple mantissas under one shared exponent, MiX groups private, per-element exponents under a single shared mantissa. To handle asymmetric VLM outlier
topologies, we introduce an adaptive dual-format (MiX-MX) inference framework. By algebraically factoring out the shared MiX mantissa, this framework maps to a custom accelerator, replacing multipliers with efficient shifters. Evaluated end-to-end on multiple VLMs, our 4.5-bit MiX formulation exhibits equivalent or superior accuracy on multi-modal
benchmarks compared to NVFP4. Simultaneously, the MiX accelerator delivers a 25% improvement in area efficiency over the NVFP4 baseline and a 2.3-4.5x speedup with 1.4-2.9x energy reduction across models compared to the state-of-the-art accelerator Focus, proving the inverted-scaling datapath is physically superior for efficient VLM deployment.
Read the full abstract ▾
Systems and Control
Bifurcation Beyond Surface-Tangential Asymptotic Convergence in Continuous Sliding Mode Control
Fan Zhang, Jingwen Xu, Peng Li et al. (+2 more)
For second-order systems under continuous sliding mode control (SMC), the literature has long relied, largely through phase-portrait illustrations, on the implicit convention that the phase-plane trajectory approaches the
equilibrium along a direction tangential to the designed sliding surface. This paper investigates this tangential convergence assumption through a rigorous phase-plane analysis of the double-integrator system subject to standard linear SMC. We reveal that the asymptotic state ratio is not unique but instead exhibits a bifurcation that depends on the control gains, the sliding surface parameter, and the initial conditions. The key finding is that the system may converge along an implicit secondary
manifold rather than aligning with the designed sliding surface, implying that smooth entry is not globally guaranteed. We derive closed-form expressions for the convergence ratios and establish a classification framework that precisely characterizes when the smooth-entry assumption holds and when it fails. The analysis is further extended to classical PD control, and we show that the bifurcation threshold coincides with the critical damping boundary that separates the two convergence regimes. These findings bridge terminal geometry and convergence smoothness, providing a predictive framework for high-performance motion control design. Simulation results validate the proposed classification of convergence regimes.
Read the full abstract ▾
All authors Fan Zhang, Jingwen Xu, Peng Li, Jun Zhou, Yaohua Guo
Robotics
VAST: V2X/Dynamic Map-Aware Autonomous Driving Systems Validation Toolchain
Shunsuke Ito and Takuya Azumi
Cooperative autonomous driving in the IoT-to-Edge-to-Cloud continuum requires system-level validation across vehicles, infrastructure sensors, edge-side Dynamic Map services, and in-vehicle autonomous-driving stacks. This paper presents VAST, a V2X/Dynamic Map-aware validation toolchain that connects Scenic, Scenario Simulator v2, AWSIM, Autoware, and SIM-LDM. VAST does not introduce a new search
algorithm; instead, it addresses interoperability challenges, including Lanelet2-to-Scenic mapping, ROS 2-based co-simulation through SS2, Dynamic Map object injection into Autoware, and collection of TTC, PET, collision, timeout, and performance measurements. In occluded-intersection scenarios, Lanelet2-compatible constrained sampling increases the edge-case discovery rate from 40.0% to 80.0% and reduces the average time per discovered edge case from 259.7 s to 110.4 s. Under the same generated scenario distribution, Dynamic Map availability reduces the collision rate from 78.0% to 40.0% and increases non-collision outcomes from 22.0% to 60.0%, with statistically significant TTC/PET shifts. A throughput study with 1-16 NPCs shows that sampling remains below 0.1 s, whereas AWSIM/Autoware execution and restart overhead dominate runtime. These results position VAST as a practical validation infrastructure for cooperative autonomous-driving CPSs.
Read the full abstract ▾
Artificial Intelligence
FINSKILLOPS: A Self-Evolving Multi-Agent System for SEC Filing QA
Yanzhang Ma, Zhenghan Tai, Hanwei Wu et al. (+25 more)
Financial QA systems are typically improved before deployment through better
retrieval, prompting, or
agent coordination, leaving their reliability behavior fixed thereafter. In practice, new SEC-filing questions repeatedly expose heterogeneous errors in period, entity, evidence use, and calculation. Existing self-improvement methods can turn failures into new behaviors, but offer limited control over where a correction should apply or which previously correct answers it may break. We therefore frame post-deployment improvement as controlled behavioral maintenance: recurring failures should become scoped skill patches, and each patch should earn deployment with- out introducing regressions. We instantiate this view in FINSKILLOPS, a multi-
agent system for SEC filing QA. FINSKILLOPS derives reusable skills from evidence-grounded, typed failure diagnoses and governs them through targeted validation, protected-case regression checks, negative controls, and versioned replacement or retirement. Across six financial QA
benchmarks, a single frozen skill registry achieves the highest verdict-weighted correctness and reference consistency among the evaluated systems. Evolved skills raise correctness from 3.70 to 4.55 on our enhanced
benchmark. In a separate 12-round operational study, only six of 33 proposed skills are promoted, while the monitoring non-correct rate falls from 20.0% to 12.5%. These results establish controlled skill scope, admission, and lifecycle management as the foundation for reliable self-improvement.
Read the full abstract ▾
All authors Yanzhang Ma, Zhenghan Tai, Hanwei Wu, Sizhe Guan, Jianliang Lei, Hailin He, Chaolong Jiang, Jijun Chi, Tung Sum Thomas Kwok, Bohuai Xiao, Jingrui Tian, Xinlu Wu, Xingao Zhan, Peng Lu, Muzhi Li, Yihong Wu, Liheng Ma, Sicheng Lyu, Tianshuo Yan, Junhao Zhu, Yaqian Xu, Lei Ding, Yufei Cui, Ziquan Liu, Boyu Han, Hengli Liu, Ling Zhou, Xinyu Wang
Game Theory
A Logarithmic Regret Bound for Optimistic Hedge in General-Sum Games
Junsoo Ha
Can simple no-regret dynamics attain smaller regret in self-play than against arbitrary adversaries? In $n$-player general-sum games, Daskalakis et al. 2021 proved an $O(n\log d_i\log^4 T)$ individual regret bound for Optimistic Hedge, which improves upon the classical $O(\sqrt T)$
adversarial regret bound. In this work, we show that Optimistic Hedge with a constant step size can further achieve $O(\sqrt n\log d_i\log T)$ individual external regret under expected loss-vector feedback. The time-averaged play consequently enjoys a coarse correlated
equilibrium gap $O(\sqrt n\log d\log T/T)$, where $d=\max_i d_i$. The improvement comes from a larger admissible step size $\eta=\Theta(1/(\sqrt n\log T))$. Our analysis proves factorial bounds on high-order differences of probability-weighted pairwise loss gaps, then applies finite-difference interpolation in a fixed Euclidean norm. These estimates sharpen the analysis of Daskalakis et al. 2021 and yield a logarithmic regret bound.
Read the full abstract ▾
Machine Learning
Conservation Buys Stability and Factoring Buys Counterfactuals in Physical World Models
Yufeng Wang, Parivesh Priye, Lu Wei et al. (+1 more)
A learned simulator can reproduce its training conditions accurately yet fail in two distinct ways once those conditions change. Over long rollouts, small errors accumulate until the trajectory drifts away from physically plausible behavior; under an intervention on a physical parameter, the model may continue to follow the law seen during training rather than the intervened one. We show that these two failures require different structural remedies. Evolving a learned energy with a
symplectic integrator preserves the geometry of the conservative dynamics and keeps rollouts bounded and physically meaningful for up to $100\times$ the training horizon, while equal-capacity predictors, an energy-regularized predictor, and a tuned neural ODE diverge. By contrast, encoding the physical coupling through an explicit linear factorization enables the model to follow a never-seen sign of that coupling, whereas an unrestricted parameterization remains locked to the training law. Crucially, the two mechanisms are separable: removing the structure responsible for long-horizon stability leaves counterfactual transfer intact, while removing the factorized coupling destroys counterfactual transfer without eliminating stability. This double dissociation, established with matched controls that remove or replace one structural component at a time, persists beyond the headline three-body system and remains visible when the physical state must be inferred from pixels rather than provided directly. The result is a concrete design principle for physical world models: long-horizon stability and changed-law generalization arise from distinct structural commitments, and each can be imposed deliberately without requiring the other.
Read the full abstract ▾
All authors Yufeng Wang, Parivesh Priye, Lu Wei, Haibin Ling
Artificial Intelligence
When2Think: Learning When and How Much to Reason
Jaejun Shim, HyunJin Kim, Young Jin Kim et al. (+1 more)
Large
Reasoning Models (LRMs) often overthink easy problems and underthink hard ones, leading to inefficient computation allocation. Existing methods regulate generated computation or select between direct answering and explicit
reasoning, but do not jointly control whether}to reason and how much computation to allocate within
reasoning. We call the resulting difficulty-dependent loss in accuracy under computation reduction the efficiency tax. We propose When2Think, an RLVR-based post-training framework for instance-adaptive computation allocation. Its core mechanism, Instance-level Difficulty-Aware Control (IDAC), uses cached reference statistics of success and
token cost to modulate a correctness-gated efficiency bonus based on generated
token count. Importance sampling supports exploration of Think and NoThink, while Batch-Wise Standardization constructs standardized advantages for critic-free optimization. The framework requires neither a learned reward model nor a learned critic, and offline reference caching avoids online reference-model queries during policy updates. On AIME24, When2Think improves Pass@3 by 10.0 percentage points while reducing
token usage by 27.9% relative to the backbone.
Read the full abstract ▾
All authors Jaejun Shim, HyunJin Kim, Young Jin Kim, JinYeong Bak
Machine Learning
CoRe: Coherence and Relational Alignment for Multivariate Time Series Forecasting
Xiaoyu Lin, Huiran Duan, Yining Liu et al. (+3 more)
Direct
forecasting has become a standard paradigm for multivariate time-series
forecasting because it predicts the full future horizon in a single pass. However, its training objective is often still decomposed into pointwise errors such as MSE. Such objectives provide stable supervision, but they do not explicitly preserve the structure of the future trajectory: temporal coherence within each variable and relational consistency across variables can both be weakened. We propose CoRe, a model-agnostic learning objective for direct multivariate
forecasting. CoRe replaces pointwise supervision with two output-space constraints: a frequency coherence loss that aligns predicted and target spectra, and a low-rank relational graph loss that matches sampled pairwise differences in a target-derived PCA subspace. The resulting objective introduces no trainable parameters and can be applied to existing
forecasting backbones by changing only the loss. Experiments on standard
benchmarks show that CoRe improves strong baselines, compares favorably with recent
forecasting objectives, and remains effective across different backbones,
datasets, and hyperparameter settings overall consistently.
Read the full abstract ▾
All authors Xiaoyu Lin, Huiran Duan, Yining Liu, Zhixiang Wu, Chu Lin, Lin Lu
Computer Vision
Beyond Patch Removal: Persistent Adversarial Effects in Vision-Language-Action Policies
Enhao Wu, Fusen Guo, Yuxin Cao et al. (+3 more)
Adversarial patches to
Vision-
Language-Action (VLA) policies can cause both immediate action corruption and persistent state effects that remain after the patch is removed. Existing evaluations largely focus on continuous attacks and do not separate these two effects. We introduce a state-restoration protocol that removes the patch at matched action-chunk boundaries and measures subsequent recoverability under the same remaining step budget. Clean, random-patch, deviation-matched, and fixed-direction controls distinguish
adversarial effects from occlusion, action-error magnitude, and directional persistence. We also evaluate a recovery adapter trained on attack-induced states under controlled intervention latency. On OpenVLA-OFT with EDPA attacks, only 36.2% of LIBERO-Long episodes remain recoverable after five chunks, compared with 89.9% and 87.0% for the deviation-matched and fixed-direction controls. Similar persistent effects are observed on autoregressive OpenVLA. The recovery adapter improves recovery from 7.7% to 47.4% at one-chunk latency, but its benefit decreases substantially with delayed intervention. These results show that
adversarial effects can persist after patch removal and that timely intervention is critical for recovery.
Read the full abstract ▾
All authors Enhao Wu, Fusen Guo, Yuxin Cao, Ziyang Lyu, Lin Li, Wei Song
Robotics
Towards High-DoF Dexterous Manipulation through VLA Post-Training
Junlei Zhu, Shenzhe Yao, Chaogui Huang et al. (+6 more)
Imitation-learned
vision–
language–action (VLA) foundation models acquire broad manipulation capabilities by scaling
robot data across tasks and embodiments, but reliable deployment on a specific downstream task and hardware platform still requires post-training. Dexterous hands make this adaptation particularly difficult: their broad behavioural repertoire and high degree of freedom create a large and structured action space. Three obstacles are central: open-source VLAs do not natively provide an action interface for high-DoF hands; gesture mismatch during human-gated DAgger takeover creates command discontinuities and contaminates corrective trajectories; and reinforcement learning in the raw joint space is sample-inefficient. We present a unified four-step post-training pipeline comprising a learned temporal hand-action codec, supervised fine-tuning, DAgger, and real-world residual reinforcement learning. The codec adapts a pretrained VLA to absolute dexterous-hand commands. Buffered rollback, pose
alignment, and smooth command blending enable continuous, task-relevant DAgger corrections, while latent residual RL confines exploration to coordinated hand motions captured by the codec. We evaluate the pipeline on five diverse real-world tasks spanning bimanual transfer, in-hand reorientation, and tool use. Within the reported post-training budgets, the resulting policies achieve 100% success on every evaluated task over 20 trials per task. These results provide a practical path for adapting VLA foundation models to reliable real-world dexterous manipulation.
Read the full abstract ▾
All authors Junlei Zhu, Shenzhe Yao, Chaogui Huang, Wenkai Zhu, Jingwei Peng, Guanqi He, Soren Schwertfeger, Jiahao Chen, Yide Liu