3769 papers
This paper challenges the assumption that LLM deception is coextensive with lying by demonstrating that models can successfully deceive through misleading non-falsities that current truth probes fail to detect, thereby revealing a critical blind spot in mechanistic deception detection.
The GATech team's approach to the AbjadGenEval shared task utilized a fine-tuned multilingual E5-large encoder with simple mean pooling to achieve an F1 score of 0.75 for detecting AI-generated Arabic text, finding that this stable baseline outperformed complex pooling strategies likely due to data limitations and a distinct length difference between human-written and machine-generated texts.
This paper demonstrates that fine-tuned bidirectional encoders, specifically a hybrid AraBERTv2 architecture, significantly outperform large-scale causal decoders in the challenging task of 82-class Arabic medical text classification by better capturing global semantic context despite data imbalance and label noise.
This paper introduces Personalized Group Relative Policy Optimization (P-GRPO), a novel framework that improves alignment with diverse individual preferences by decoupling advantage estimation from batch statistics and normalizing rewards against preference-group-specific histories, thereby overcoming the limitations of standard GRPO in handling heterogeneous user signals.
This paper addresses the regulatory ambiguity surrounding "AI models" and "AI systems" by proposing clear conceptual and operational definitions that distinguish trained parameters from broader system components, thereby facilitating the precise allocation of obligations across the AI value chain.
LWM-Temporal is a task-agnostic foundation model for wireless channels that leverages a novel Sparse Spatio-Temporal Attention mechanism and physics-informed pretraining to learn universal, geometry-consistent embeddings, achieving superior performance in channel prediction across diverse mobility regimes with limited fine-tuning data.
HTM-EAR is a hierarchical tiered memory system that combines HNSW-based working memory with archival storage, importance-aware eviction, and hybrid routing to effectively preserve essential information and maintain high retrieval precision under sustained saturation, significantly outperforming traditional LRU approaches while approaching the performance of unbounded oracle memory.
Tureis is a self-supervised, Transformer-based framework designed for edge deployment in smart homes that achieves fine-grained, multi-failure sensor localization by encoding heterogeneous sensor streams into compact features and using a masked reconstruction objective to identify faulty devices without requiring labeled data or human intervention.
This paper evaluates how autonomous cyber attack agents generalize to unseen IP reassignments in enterprise networks, finding that while prompt-driven LLM agents achieve the highest success rates compared to traditional RL and adaptation methods, they incur significant costs in computational efficiency, transparency, and reliability due to issues like repetition loops.
This large-scale controlled study reveals that evaluation format (multiple-choice vs. open-ended) and specific model-scaffold interactions, rather than scaffold architecture alone, are the primary drivers of measured safety differences in language models, ultimately demonstrating that no universal safety ranking exists across different deployment configurations.
This paper proposes a parameter-efficient continual learning framework for Human Activity Recognition that mitigates catastrophic forgetting in domain-incremental scenarios by employing channel-wise gated modulation to adapt frozen pretrained representations through bounded diagonal scaling, thereby achieving superior stability and plasticity with minimal parameter updates.
This paper introduces eXplicit Sharpness-Aware Minimization (XSAM), a novel and computationally efficient framework that improves upon existing SAM implementations by providing a more faithful interpretation of the ascent gradient and explicitly estimating the direction toward local loss maxima to achieve superior generalization.
This paper introduces \texttt{InFusionLayer}, a novel, open-source Python ensemble tool inspired by Combinatorial Fusion Analysis (CFA) that integrates rank-score characteristic functions and cognitive diversity to enhance multiclassification performance across PyTorch, TensorFlow, and Scikit-learn workflows.
This paper addresses the failure of byte-sequence-based masked modeling in encrypted traffic classification by identifying a mismatch in inductive bias and proposing FlowSem-MAE, a protocol-native tabular masked autoencoder that leverages field-specific semantics and temporal patterns to significantly outperform existing methods with substantially less labeled data.
OmniGuide is a universal framework that enhances the performance of generalist Vision-Language-Action (VLA) policies on complex tasks by integrating diverse guidance sources—such as 3D foundation models and human pose estimators—into a unified differentiable energy field that guides action sampling through task-specific attractors and repellers.
This paper proposes CAADRL, a cluster-aware deep reinforcement learning framework that leverages hierarchical encoding and a dynamic dual-decoder to efficiently solve Pickup and Delivery Problems by explicitly modeling multi-scale cluster structures, achieving state-of-the-art performance with significantly lower inference latency than collaborative-search baselines.
This paper resolves a 20-year-old conjecture by proving that the volume-averaged Ricci scalar of binary Bayesian networks is universally quantized to positive half-integers for tree and complete-graph structures via a Beta function cancellation mechanism, while demonstrating that this quantization fails in general due to loop counterexamples and contrasting the positive curvature of discrete networks with the negative curvature of Gaussian DAGs.
This paper proposes using Neural Cellular Automata to generate controllable, synthetic non-linguistic data for pre-pre-training large language models, demonstrating that this approach significantly improves downstream performance and convergence speed while outperforming training on much larger natural language datasets.
This paper introduces HTMuon, a heavy-tailed spectral correction method that improves upon the Muon optimizer by preserving parameter interdependencies while inducing heavier-tailed weight spectra, resulting in consistent performance gains in LLM pretraining and image classification alongside theoretical convergence guarantees.
This paper introduces Search Agent Policy Optimization (SAPO), a method that resolves catastrophic training instability in Tool-based Agentic Reinforcement Learning by applying a conditional token-level KL constraint to prevent Importance Sampling Distribution Drift, achieving significant performance gains with only a single line of code modification to standard GRPO.