LLMs & Texto — Radar de IA

Don't Mix Rewards, Mix Policies: Policy Decomposition and Optimization for Multi-Reward RL

arXiv:2607.29246v1 Announce Type: new Abstract: Modern large language models (LLMs) are expected not just to answer correctly, but to adapt their behavior to different human values and use cases. As a result, multi-reward reinforcement learning (RL) has become an increasingly important problem for LLMs, where each reward captures a different aspect of desired behavior. However, optimizing with multiple rewards suffers from a more severe alignment tax issue, where different optimization objective...

03.08.2026

Blog LLMs & Texto

Safety, or Just Capability? A Validity Audit of Agent-Safety Benchmarks

arXiv:2607.28685v1 Announce Type: new Abstract: Agent-safety benchmarks measure different behaviors, and their scores get quoted interchangeably as an agent's safety. We treat four of them (R-Judge, InjecAgent, AgentHarm, AgentDojo) as measurements to be validated, running each under its official implementation and author-provided scorer on up to 22 models, with MMLU and GPQA measured by us under one protocol as a capability composite. The metric is the first problem. On any binary trace-judgmen...

03.08.2026

Blog LLMs & Texto

WaiT for the Signal: Simple Frequency-Aware Flow-Matching

arXiv:2607.28760v1 Announce Type: new Abstract: As image generation models scale to ever higher resolutions, global coherence, local detail, and texture fidelity become critical axes for generation quality. However, standard flow matching treats all spatial frequencies uniformly, ignoring the natural frequency hierarchy where high-frequency bands become indistinguishable from pure noise far earlier than coarse structures. We introduce WaiT, a Wavelet-aware image Transformer that decomposes gener...

03.08.2026

Blog LLMs & Texto

A Generalized-Bayes Perspective on Counterfactual Explanations: Posterior-Based Decision-Making and Evaluation

arXiv:2607.29077v1 Announce Type: new Abstract: Counterfactual explanations (CEs) enhance the interpretability of machine learning models by identifying the smallest change to an input required to obtain a desired output. Although CEs are conventionally formulated as a distance-minimization problem, the theoretical basis of this formulation has received limited attention. We show that a distance-minimization-based CE is mathematically equivalent to the maximum a posteriori (MAP) estimate of a Gi...

03.08.2026

Blog LLMs & Texto

CAGE: Certified Authorization under Typed-Return Uncertainty for Tool-Using Agents

arXiv:2607.29190v1 Announce Type: new Abstract: Tool-using LLM agents act on typed tool returns, records pairing provenance and categorical fields with numerical values. Runtime permission gates generally authorize the observed return and action, leaving the decision unprotected against small errors in how the return was bound to its source. We ask whether a candidate action stays authorized over a declared neighborhood of plausible correctly bound returns: one admissible binding fault plus boun...

03.08.2026

Blog LLMs & Texto

Harnessing the Wisdom of LLM Crowds through Complementarity-Driven Iterative Collaboration

arXiv:2607.29087v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in enterprise settings, yet individual models remain bounded by model-specific capability limitations. These heterogeneous boundaries pose a deployment challenge, but also create an opportunity: strategically coordinating multiple LLMs may unlock collective intelligence exceeding any single model. Existing approaches fix how models are combined in advance, overlooking the dynamic, state-depende...

03.08.2026

Blog LLMs & Texto

Outcome-Guided Distillation: A Teacher-Student Framework to Advance VLM Reasoning in Autonomous Driving

arXiv:2607.29052v1 Announce Type: new Abstract: End-to-end (E2E) autonomous driving aims to learn a direct mapping from visual observations to control actions. However, these E2E models often act as black boxes and struggle with complex scenarios. To address this, recent works incorporate Vision-Language Models (VLMs) to provide explicit reasoning, enhancing both interpretability and driving robustness. These approaches typically rely on pre-generated annotations, which suffer from potentially f...

03.08.2026

Blog LLMs & Texto

Fragility of Value under Imperfect Alignment

arXiv:2607.28881v1 Announce Type: new Abstract: As more responsibility is placed upon AI systems, it becomes increasingly important to guarantee that these systems are aligned with humanity. A common fear in AI safety is that human value is fragile -- that is, optimizing too heavily for an imperfect proxy to human values will lead to a catastrophic outcome. In this paper, we present a model of the alignment problem where an agent undergoes idealized alignment training that guarantees its value f...

03.08.2026

Blog LLMs & Texto

Evidence-Grounded Constraint Checking in Construction Documents

arXiv:2607.29058v1 Announce Type: new Abstract: Professional-document review is a constraint-checking problem in which decisions depend on relations among text, geometry, pages, and document revisions. We present an evidence-grounded pipeline that normalizes extracted facts, executes four-state rules deterministically, retains source spans, and escalates unresolved cases. We evaluate its PDF evidence allocator on 160 reference-based tasks from 29 construction projects using a repeated four-syste...

03.08.2026

Blog Robótica & RL

Auto-JEPA: A Latent World Model of Continuous Intent for End-to-End Autonomous Driving

arXiv:2607.29031v1 Announce Type: new Abstract: Existing autonomous-driving world models typically perform dense prediction of future videos, occupancy states, BEV representations, or agent motion. We argue that planning need not reconstruct the complete future world, but only focus on scene features that affect future ego action. Based on this perspective, we propose Auto-JEPA, an action-oriented latent world model that learns continuous future driving intent through joint-embedding prediction....

03.08.2026

Blog LLMs & Texto

ReMoE: Report-Guided Mixture-of-Experts for Multimodal OCT/OCTA Anomaly Detection

arXiv:2607.29039v1 Announce Type: new Abstract: Multimodal medical anomaly detection identifies samples deviating from normal patterns, where scarce abnormal cases make normality modeling from normal data practical. In retinal Optical Coherence Tomography (OCT) and OCT Angiography (OCTA) anomaly detection, existing unsupervised methods rely on visual feature distributions, reconstruction residuals, or encoder-decoder discrepancies, making anomaly scores depend on appearance-level deviations, whi...

03.08.2026

Blog Multimodal

Mitigating Class-Tail Undercoverage in Medical Vision-Language Models under Clinical Shift

arXiv:2607.28696v1 Announce Type: new Abstract: Medical vision-language models (VLMs) can retain high observed marginal coverage after clinical shift while substantially under-covering an individual disease class. The affected class varies with acquisition protocol and backbone geometry, so source prevalence does not reliably reveal the failure. Existing localized and tail-aware conformal methods respectively adapt to test neighborhoods and source-frequency tails, leaving held-out class-wise cov...

03.08.2026