Noticias de IA - ítems

arXiv:2610.33767v1 Announce Type: new Abstract: Zero-shot detection of machine-generated text is vital for mitigating the risks of large language models (LLMs) misuse. While the state-of-the-art dual-model detectors and resampling-based detectors outperform single-model statistical detectors on unprompted continuations, they suffer a catastrophic breakdown on instruction-conditioned text in the black-box setting, where the source model is unknown. We analyze this failure and trace it to poor prompt conditioning for resampling-based methods and diminished cross-model discrepancy for dual-model detectors. Motivated by recent studies that human writing maintains a steady, uniform information flow, whereas machine-generated text exhibits sharp, localized entropy fluctuations, we propose a new metric, the Relative Peak Prominence Rate (RelPPR) to capture this structural contrast along the entropy trajectory and further leverage it as a signal for LLM-generated text detection. Extensive evaluations across diverse model families and datasets demonstrate that RelPPR achieves state-of-the-art black-box detection performance and remains the most reliable detector under various practical situations. Our code is available at https://github.com/brushing-git/entropy-llm-detector.

arXiv:2610.33764v1 Announce Type: new Abstract: Multimodal reasoning requires models to combine multiple pieces of evidence into decisions that satisfy all relevant constraints. Evaluations centered on final-answer accuracy leave unclear whether failures arise from reading individual facts or using them together. We study this question through the multi-evidence integration gap, linking local fact recovery to global decision validity on controlled subset-selection and path-finding tasks. Our evaluation combines conditional performance given correct local answers with same-grid local queries, text-oracle inputs, retrieval controls, and constraint ablations. We further examine hidden-state probes and activation replacement across network depth, token positions, and model families. Across the evaluated models, accurate local recovery coexists with substantial global failure. For example, Gemma-3-27B recovers all evaluated isolated item facts correctly but achieves only 19.75% global validity on eight-item visual instances. Probes decode information about solution membership from global-prompt states, while corresponding clean-state replacement recovers failed decisions over model-dependent depth ranges. These findings show that recovering facts and representing solution-related information are insufficient for reliable joint decisions. They motivate evaluations and model improvements that explicitly target how available evidence is combined, alongside advances in perception and model scale.

arXiv:2610.33726v1 Announce Type: new Abstract: Modeling human driving behavior is fundamental to understanding driver decision-making and supporting autonomous vehicles (AVs) in mixed traffic. Reward learning offers a principled way to infer the latent objectives behind driving behaviors, yet most existing methods are formulated in single-agent settings and treat surrounding vehicles as dynamic obstacles. Directly extending reward learning to multi-agent scenarios is difficult because the joint action space grows exponentially with the number of interacting vehicles. In addition, existing approaches mainly rely on physical demonstrations and rarely exploit human feedback during reward recovery. To address these limitations, this paper proposes a Human-Feedback-Guided Multi-Agent Inverse Reinforcement Learning (HFG-MAIRL) framework for interactive driving behavior modeling. First, we develop a multi-agent maximum-entropy inverse reinforcement learning module that explicitly models each vehicle and its directly interacting neighbors in a localized Markov game, while using aggregated traffic states to capture the influence of the broader traffic environment. The resulting coupled multi-vehicle responses are approximated through finite marginal-response iterations over factorized vehicle-level policies, avoiding explicit enumeration of the full joint action space. Second, we incorporate human feedback as ordinal preference supervision through a Bradley-Terry likelihood. In our experiments, driver emotion labels are used as one instantiation of such feedback and converted into pairwise ordinal preference constraints. Experiments on the AIDE dataset show that HFG-MAIRL outperforms representative IRL baselines in decision modeling and produces interpretable feedback-aligned reward patterns.

arXiv:2610.33702v1 Announce Type: new Abstract: Typed decision-making tasks in agents, such as UI element selection, product recommendation, and chart matching, require models to choose from a set of discrete candidates. While prior work demonstrates that text-only System One models achieve four key properties including candidate-order invariance, scalability, and cache reuse via separate request-candidate encoding, extending them to multimodal settings where candidates are visual remains underexplored. We introduce Iris, a framework that accommodates image, text, or hybrid candidates while preserving the core decision module. We prove that order invariance and exact cache reuse hold for any per-candidate encoder under Iris, and derive the closed-form latency reduction. We construct a visual-candidate benchmark from Multimodal-Mind2Web (K=16, deduplicated, order-shuffled, three generalization splits). Independent encoding (Iris) achieves 78.4 / 72.1 / 74.8 percent accuracy with zero-flip rate under candidate reordering. Joint encoding reaches 90.6 / 89.5 / 91.6 percent but suffers an 8 to 10 percent answer-flip rate when shuffled. To bridge this gap, we propose a design that encodes each candidate conditioned on the request state without inter-candidate cross-attention; it preserves strict order invariance, enables forked cache sharing, and attains 91.7 / 86.5 / 90.0 percent accuracy at one-third of OneJev latency with calibrated null probabilities (AUROC 0.93 to 0.97). Reusing image candidates across requests cuts latency by 37x at K=256 and 191x at K=2048. On MetaEncoder benchmarks, Iris-4B achieves 100 percent on JevBench original and easy, outperforming 30B MetaEncoder and 27B OpenJev. Finally, we resolve a systems bottleneck on Ampere GPUs: reformulating Qwen3.5 patch-embedding Conv3d into its exact matrix-product form accelerates processing by 350x. Code, data, and checkpoints: https://github.com/lllfx/Iris.

arXiv:2610.33690v1 Announce Type: new Abstract: Many mathematical research agents focus on solving problems with fixed statements. However, conjecture research, especially in applied mathematics, may require revising candidate questions or conclusions as investigation proceeds, coupling problem formulation with proof development. We present PACE, a research-proof agent that supports this process through a structured research loop that selects mathematical objectives and develops or challenges candidate mechanisms. The whole prover constructs proofs from these mechanisms, and an independent verifier checks them against the original target. Evidence-tracking memory stores assumptions and dependencies arising during research and marks evidence for revalidation when a candidate or condition changes. This memory supports task-specific context composition that keeps checked premises, exploratory material and unresolved obligations distinct. To evaluate proof construction and conjecture research, we introduce ResearchProofBench, an applied-mathematics benchmark focused on optimization, comprising 51 known-proof tasks with 331 hidden checkpoints under historical abstract access and six open-problem cases. The known-proof tasks assess reconstruction of published results, while the open-problem cases assess mathematical investigation and proof development. On the 51 known-proof tasks, PACE achieves an aggregate Proof Score of 75.16 and completes 20 proofs. In conjecture research, PACE develops conditional results, with its submissions receiving higher review scores than the compared baselines.

arXiv:2610.33684v1 Announce Type: new Abstract: Self-evolving training improves the performance of agents on difficult tasks through verifiable feedback, but it often applies costly deliberation uniformly across tasks. To address this problem, we propose Mode-Optimized Reasoning with Self-Adaptive Effort (MORE), a self-evolving training framework that adaptively selects both reasoning form and computational effort for each task. Specifically, MORE follows a two-stage training paradigm: (i) \emph{Stage 1} establishes a compact prior over direct answering, short and long chain-of-thought, and code-based reasoning, and (ii) \emph{Stage 2} self-evolves this prior with Mode-Adaptive and Test-time Effort PPO (MATE-PPO) on generated, verifiable task streams without curated data. By learning mode selection and token expenditure jointly, MORE allocates concise reasoning to tasks that permit it while retaining extended deliberation where it is useful. Across coding, mathematics, and general-reasoning benchmarks, MORE achieves consistent performance gains on Base and Coder backbones at model scales from 3B to 14B. During inference, MORE demonstrates higher accuracy and better efficiency, with markedly shorter response lengths and reduced inference time.

arXiv:2610.33674v1 Announce Type: new Abstract: Medical vision--language pretraining learns transferable representations from paired images and radiology reports, yet differences in scanners and acquisition protocols can impair transfer. Existing approaches primarily enrich clinical semantics, leaving structured acquisition information in DICOM headers underused. We introduce Meta-CLIP, a metadata-conditioned contrastive framework that uses acquisition similarity to reweight cross-modal negatives during pretraining. A kernel over numerical and categorical DICOM fields defines a bounded pairwise margin. This margin increases the contribution of acquisition-similar negatives, emphasizing distinctions among studies that share imaging conditions. The kernel remains fixed during image--text optimization, and inference follows the standard dual-encoder architecture without requiring metadata. We evaluate Meta-CLIP on chest radiography and volumetric chest CT through zero-shot classification, appearance perturbations, grounding, and representation analyses. On CT-RATE, Meta-CLIP improves macro-AUROC over CT-CLIP by 1.41 points, with a 95\% paired bootstrap confidence interval of 0.96--1.85 points. It also achieves the highest macro-AUROC on four of five CXR benchmarks. An exploratory BrgSA evaluation extends the analysis to another CT backbone. These findings support acquisition metadata as a practical training-time prior for medical vision--language learning. Code will be made publicly available upon publication.

arXiv:2610.33667v1 Announce Type: new Abstract: Resolving a repository-level issue typically requires a software-engineering (SWE) agent to localize the faulty code, reproduce the failure, and patch the code. Reinforcement learning trains such agents from a single test result on the final patch, while on-policy self-distillation (OPSD) adds token-level supervision from a teacher with privileged evidence, exposing the full reference patch at every step. Neither provides stage-specific guidance for localization or reproduction. We propose SWE-Ladder, which turns the verified repair of each training instance into stage-specific evidence: the changed files and lines for localization, the fail-to-pass tests for reproduction, and the diff for patching. A rule-based router assigns each step of an end-to-end rollout to a stage based on its tool call, a frozen teacher sees only the matching evidence, and the stages are distilled in repair order while GRPO optimizes the full trajectory. With Qwen3-4B-Instruct-2507 on SWE-bench Verified under the R2E-Gym scaffold, SWE-Ladder improves pass@1 from 9.6% to 16.4%, compared with 10.4% for GRPO and 12.4% for OPSD with the full patch. Under the OpenHands scaffold, it improves pass@1 from 13.0% to 17.6%. Reversing the stage order under R2E-Gym reduces pass@1 to 11.4%.

arXiv:2610.33658v1 Announce Type: new Abstract: Vision-language-action (VLA) models have demonstrated strong capabilities in robotic manipulation, yet their reliance on dense, continuous-valued computation presents a fundamental challenge for energy-constrained deployment. Inspired by the sparse, event-driven signaling and temporal integration mechanisms of biological neurons, a promising alternative is to introduce spiking computation into VLA models. However, enabling such computation without restructuring pretrained policies or compromising closed-loop control remains an open challenge. To address this challenge, we introduce the Power-of-Two Leaky Integrate-and-Fire (PoT-LIF) neuron, which emits signed spikes with power-of-two amplitudes. By combining magnitude-coded spike events with residual-preserving membrane dynamics, PoT-LIF enables continuous information to be progressively represented through discrete spikes. The Heartbeat formulation further localizes this process within weighted projections, bridging biologically inspired spiking computation and pretrained VLA architectures without modifying their external inference interfaces. We integrate PoT-LIF into the vision projection, vision-language projection, and action projection of $\pi_{0.5}$ and evaluate both operating modes over 2,000 paired closed-loop LIBERO episodes per configuration. Standard-mode PoT-LIF VLA achieves success rates of 96.90% and 96.35% at $T = 1$ and $T = 2$, respectively, while Heartbeat mode achieves 97.25% and 96.90% at $T = 2$ and $T = 4$, compared with 96.70% for the original continuous-activation $\pi_{0.5}$ baseline. At $T = 1$ with a 60.01% firing rate, PoT-LIF uses approximately 1/7.67 of the dense-reference arithmetic energy, corresponding to a modeled reduction of 86.95%.

arXiv:2610.33654v1 Announce Type: new Abstract: High penetrations of distributed energy resources (DERs) and fast-varying electric vehicle (EV) charging intensify voltage and line-limit violations in distribution networks (DNs), making safe, low-cost operation increasingly challenging. This paper proposes a physics-informed safe deep reinforcement learning (DRL) framework for coordinating community battery energy storage systems to mitigate constraint violations while minimizing operating cost. To represent nonlocal spatiotemporal interactions, we develop a dual-level physics-informed graph representation in which line impedance and capacity shape spatial information propagation, while power-balance, voltage-profile, and flow-reconstruction losses regularize the latent state toward control-relevant physical consistency. Building on this representation, we develop Lagrangian Dedicated-Critic PPO (LDCPPO), which implements constraint-specific value decomposition: independent reward and cost critics preserve heterogeneous voltage- and line-loading-related learning signals and combine their advantages only during the Lagrangian policy update. Case studies on DNs with high rooftop photovoltaic penetration and EV charging demonstrate that the proposed method reduces the key voltage-violation degree by up to 80.5% compared with the tested baselines, while also achieving the lowest voltage-violation count and line-overload degree with competitive operating cost.