arXiv:2603.12269v1 Announce Type: new Abstract: Early-exit deep neural networks enable adaptive inference by terminating computation when sufficient confidence is achieved, reducing cost for edge AI accelerators in resource-constrained settings. Existing methods, however, rely on suboptimal exit policies, ignore input difficulty, and optimize thresholds independently. This paper introduces DART (Input-Difficulty-Aware Adaptive Threshold), a framework that overcomes these limitations. DART introduces three key innovations: (1) a lightweight difficulty estimation module that quantifies input complexity with minimal computational overhead, (2) a […]

Ver mais

arXiv:2604.08584v1 Announce Type: new Abstract: Long-context LLMs increasingly rely on extended, reusable prefill prompts for agents and domain Q&A, pushing attention and KV-cache to become the dominant decode-time bottlenecks. While sparse attention reduces computation and transfer costs, it often struggles to maintain accuracy at high sparsity levels due to the inherent distribution shift between Queries and Keys. We propose Centroid-Scoring Attention (CSAttention), a training-free sparse attention method optimized for high-throughput serving of reusable contexts. CSAttention adopts a storage-for-computation […]

Ver mais

Learning methods using synthetic data have attracted attention as an effective approach for increasing the diversity of training data while reducing collection costs, thereby improving the robustness of model discrimination. However, many existing methods improve robustness only indirectly through the diversification of training samples and do not explicitly teach the model which regions in the input space truly contribute to discrimination; consequently, the model may learn spurious correlations caused by synthesis biases and artifacts. Motivated by this limitation, […]

Ver mais

arXiv:2602.01002v1 Announce Type: new Abstract: Large language models often exhibit increased sycophantic behavior after preference-based post-training, showing a stronger tendency to affirm a user’s stated or implied belief even when this conflicts with factual accuracy or sound judgment. We present a formal analysis of how alignment from human feedback can increase this failure mode by identifying an explicit amplification mechanism that causally links optimization against a learned reward to bias in the human preference data used for alignment. […]

Ver mais

The Whole Problem Starts With This Animation The animation below is not showing fake progress. The industry has genuinely become very good at killing the stupid hallucinations: the obvious mistakes, the embarrassing answers, the screenshots everyone loves to mock. Those surface failures have been pushed back by real engineering. But that does not mean hallucinations are solved. It means the easy version is being cleaned up, while the deeper ones remain. That is what we can call the […]

Ver mais

arXiv:2602.07992v1 Announce Type: cross Abstract: The emergence of compositional reasoning in large language models through reinforcement learning with verifiable rewards (RLVR) has been a key driver of recent empirical successes. Despite this progress, it remains unclear which compositional problems are learnable in this setting using outcome-level feedback alone. In this work, we theoretically study the learnability of compositional problems in autoregressive models under RLVR training. We identify a quantity that we call the task-advantage ratio, a joint property […]

Ver mais

Direct Advantage Estimation (DAE) has been shown to improve the sample efficiency of deep reinforcement learning algorithms. However, its reliance on full environment observability limits its applicability in realistic settings, and its requirement to model transition probabilities incurs substantial computational overhead for high-dimensional observations. In the present work, we address both limitations. First, we extend the theoretical framework of DAE to partially observable domains with minimal modifications. Second, we reduce its computational complexity by introducing discrete latent dynamics […]

Ver mais

arXiv:2604.04979v1 Announce Type: new Abstract: Coding agents repeatedly consume long tool observations even though only a small fraction of each observation matters for the next step. We study task-conditioned tool-output pruning: given a focused query and one tool output, return the smallest verbatim evidence block the agent should inspect next. We introduce a benchmark of 11,477 examples built from SWE-bench repository interactions and synthetic multi-ecosystem tool outputs, with a manually curated 618-example test set. We fine-tune Qwen 3.5 […]

Ver mais

arXiv:2410.19815v2 Announce Type: replace-cross Abstract: Deep learning is advancing EEG processing for automated epileptic seizure detection and onset zone localization, yet its performance relies heavily on high-quality annotated training data. However, scalp EEG is susceptible to high noise levels, which in turn leads to imprecise annotations of the seizure timing and characteristics. This “label noise” presents a significant challenge in model training and generalization. In this paper, we introduce Bayesian UncertaiNty-aware Deep Learning (BUNDL), a novel algorithm that […]

Ver mais

This article explains the practical differences between AI workflows, autonomous agents, and multi-agent systems through real-world analogies, production trade-offs, and code examples. It argues that workflows are best for deterministic, structured tasks with predictable execution paths, while agents are better suited for open-ended problems requiring dynamic tool selection and adaptive reasoning. Multi-agent systems introduce specialized coordination between multiple agents but also increase operational complexity, debugging overhead, and cost. The piece also explores hybrid architectures, beginner mistakes, production reliability, […]

Ver mais