The deep learning revolution has enabled major AI advances in physical sciences, particularly in discovering dynamical systems from observational data. Yet the reliability of learned surrogates and symbolically discovered models is often undermined by non-uniqueness: models may fit observed trajectories perfectly, but...
Zakhar Shumaylov, P. Zaika, Philipp Scholl et al.· arXiv.org· 11 citations· ⚡1
Modern ML research increasingly faces an evaluation bottleneck: building benchmarks and running experiments is costly, slowing iteration. We ask how much can be inferred about evaluation outcomes before experiments, given only a natural-language description of the task and setup. We call this setting anticipatory evalu...
Jungsoo Park, Ethan Mendes, Gabriel Stanovsky et al.· 0 citations
Large Language Models (LLMs) with billions of parameters are prime targets for network pruning, which removes some model weights without hurting performance. Prior approaches such as magnitude pruning, SparseGPT, and Wanda either concentrated solely on weights or integrated weights with activations for sparsity. Howeve...
Rocktim Jyoti Das, Mukul Ranjan, Mingjie Sun et al.· 0 citations
We study contextual bandit problems with correlated arms and access to surrogate reward signals produced by a machine learning model, motivated by applications such as large language model (LLM) routing. Unlike classical contextual bandits that rely solely on bandit feedback and assume conditional independence across a...
Real-world time series exhibit complex dynamics characterized by multiple simultaneous temporal patterns: short-term fluctuations, periodic seasonal cycles, long-term trends, and irregular abrupt changes. However, most forecasting architectures rely on single-path temporal modeling---transformers capture long-range dep...
A comprehensive study of activation-steering-induced emergent misalignment reveals activation steering as a significant yet underexamined source of emergent misalignment and provides an activation-space perspective on emergent misalignment and its safety risks.
Qi Cao, Jian Lou, Mei-Ting Liu et al.· 0 citations
Finetuning data selection requires balancing two competing goals: selecting examples that improve the downstream objective, and doing so without repeatedly finetuning models. Train-free selectors are scalable but rely on proxies such as embedding similarity or clustering, which may not match the target objective. Train...
Ning Wang, Zhengxin Zhang, Maosen Tang et al.· 0 citations
Physics-informed neural networks (PINNs) are powerful surrogates for differential equations but are notoriously difficult to train due to spectral bias, stiffness, and poor accuracy on high-frequency or multiscale solutions. Adversarial training based on generative adversarial networks (GANs) has recently gained surpri...
Yuan-dong Cao, Chi Chiu SO, Jun-Min Wang et al.· 0 citations
Large language models (LLMs) are trained for downstream tasks by updating their parameters (e.g., via RL). However, updating model parameters forces them to absorb task-specific information, which can result in catastrophic forgetting and loss of plasticity. In contrast, in-context learning with fixed LLM parameters ca...
Rishabh Tiwari, Kusha Sareen, Lakshya A Agrawal et al.· 0 citations
Diffusion Language Models (dLLMs) have garnered significant attention for their potential in highly parallel processing. The parallel capabilities of existing dLLMs stem from the assumption of conditional independence at high confidence levels, which ensures negligible discrepancy between the marginal and joint distrib...
Haohui Zhang, Zhiye Wang, Xiaoying Gan et al.· 0 citations
Multi-agent reinforcement learning (MARL) is increasingly used to design learning-enabled agents that interact in shared environments. However, training MARL algorithms in general-sum games remains challenging: learning dynamics can become unstable, and convergence guarantees typically hold only in restricted settings...
Addison Kalanther, Sanika Bharvirkar, Shankar Sastry et al.· 0 citations
