9572 papers
This paper demonstrates that steering vectors in large language models are fundamentally non-identifiable, as numerous distinct interventions—including orthogonal perturbations—produce behaviorally indistinguishable results, thereby revealing inherent limits in interpreting these vectors as unique internal representations without additional structural constraints.
LatentChem introduces a latent reasoning interface that decouples chemical computation from textual generation, enabling models to perform multi-step reasoning in continuous latent space which spontaneously emerges as a more efficient and accurate alternative to explicit Chain-of-Thought, achieving a 59.88% win rate and 10.84 speedup over baselines.
This paper demonstrates that systematically learning embedding magnitudes by independently controlling query and document normalization significantly improves retrieval and RAG performance—particularly in out-of-domain scenarios—by revealing that magnitude encodes distinct, beneficial roles for queries and documents that are lost when assuming magnitude is noise.
This paper introduces Topology-Aware PINNs (TAPINN), a novel framework that employs supervised metric regularization and alternating optimization to effectively resolve spectral bias and mode collapse in multi-regime physics-informed neural networks, achieving superior convergence stability and accuracy compared to standard and hypernetwork-based baselines.
This paper empirically demonstrates that while Kolmogorov-Arnold Networks (KANs) can compete with MLPs on simple univariate residuals in hard-constrained recurrent physics-informed architectures, they suffer from severe hyperparameter fragility, instability in deeper configurations, and consistent failure on multiplicative terms, ultimately revealing limitations in their additive inductive bias for modeling state coupling in oscillatory systems.
This paper introduces Progressive Self-Correction (ProSeCo), a framework for Masked Diffusion Models that iteratively refines both unmasked and previously generated tokens during sampling to mitigate error accumulation, thereby achieving superior sample quality and faster inference compared to standard approaches.
The paper introduces QTabGAN, a hybrid quantum-classical generative adversarial network that leverages the expressive power of quantum circuits to synthesize realistic tabular data, demonstrating significant performance improvements over state-of-the-art models in scenarios with scarce or privacy-restricted data.
This paper introduces WeightCaster, a framework that reformulates out-of-support generalisation as a weight-space sequence modelling task to generate reliable, uncertainty-aware predictions without explicit inductive biases, demonstrating competitive performance on both synthetic and real-world datasets.
This study demonstrates that a neural network-based simulation inference framework effectively and efficiently estimates parameters in a large-scale labour market agent-based model, outperforming traditional Bayesian methods and summary statistics in recovering original parameters across various dataset scales.
This paper proposes minimax-optimal online conformal prediction algorithms for non-stationary data streams with distribution drift, utilizing drift detection to adaptively update calibration sets and leveraging model stability to achieve optimal training-conditional cumulative regret.
SubQuad is an end-to-end pipeline that overcomes the computational and data imbalance bottlenecks in large-scale adaptive immune repertoire analysis by integrating near-subquadratic MinHash retrieval, GPU-accelerated affinity kernels, and fairness-constrained clustering to enable scalable, bias-aware discovery of clinically relevant clonotypes.
This paper proposes a three-stage curriculum learning framework that leverages structure-aware masking and Group Relative Policy Optimization (GRPO) to efficiently distill Chain-of-Thought reasoning into compact student models, achieving significant accuracy gains and output length reduction on GSM8K by progressively guiding the model from structural understanding to self-optimized brevity and targeted knowledge internalization.
CityGuard is a privacy-preserving, graph-aware transformer framework that enables robust, bias-resilient person re-identification across distributed urban cameras by integrating dispersion-adaptive metric learning, spatially conditioned attention for coarse geometric alignment, and differentially private embeddings to balance retrieval accuracy with data protection.
This paper introduces RA-QA, a comprehensive benchmarking system featuring a standardized pipeline, a large-scale dataset of 9 million diverse question-answer pairs, and a unified evaluation protocol to assess and expose the limitations of respiratory audio question-answering models under real-world heterogeneity.
The paper introduces cc-Shapley, a causal-context modification of Shapley values that leverages knowledge of the data-generating process to eliminate misleading feature attributions caused by collider bias and suppression, thereby providing more accurate multivariate feature importance for explainable AI and scientific discovery.
This paper extends kernel density estimation and hierarchical shrinkage to Hoeffding trees for imbalanced regression in data streams, demonstrating that kernel density estimation significantly improves early-stream performance while hierarchical shrinkage offers limited gains.
Zatom-1 is the first open-source, end-to-end foundation model that unifies generative and predictive learning for 3D molecules and materials using a multimodal flow matching objective, achieving state-of-the-art performance across domains while significantly reducing inference time and enabling positive transfer between chemical systems.
This paper proposes a regularized online RLHF framework using Generalized Bilinear Preference Models to identify Nash Equilibria, establishing the first statistically efficient, dimension-free regret bounds for high-dimensional settings through two simple algorithms that leverage strong convexity and low-rank structures.
This paper introduces Lap2, a novel framework that enables L2-norm clipping for Laplace DP-SGD in high-dimensional models by leveraging majorization theory and Schur-convexity to overcome dimensionality barriers, thereby achieving privacy-utility performance comparable to or exceeding Gaussian DP-SGD.
This paper introduces a general inference-time optimization framework that generates experiment-grounded protein ensembles by optimizing latent representations and employing novel sampling schemes, thereby overcoming the limitations of current diffusion-based methods to produce thermodynamically plausible structures with improved agreement to experimental data while exposing vulnerabilities in existing confidence metrics.