We consider the problem of learning the optimal policy for Markov decision processes with safety constraints. We formulate the problem in a reach-avoid setup. Our goal is to design online reinforcement learning algorithms that ensure safety constraints with arbitrarily high probability during the learning phase. To this end, we first propose an algorithm based on the optimism in the face of uncertainty (OFU) principle. Based on the first algorithm, we propose our main algorithm, which utilizes entropy regularization. […]

Ver mais

I enjoyed my conversations with Timnit when she was in the MSR-NYC lab, so her situation has been on my mind throughout NeurIPS. Piecing together what happened second-hand is always tricky, but Jeff Dean’s account and Timnit’s agree on a basic outline. Timnit and others wrote a paper for FAccT which was approved for submission by the normal internal review process, then later unapproved. Timnit threatened to leave unless various details about this unapproval were clarified. Google then […]

Ver mais

arXiv:2601.05253v1 Announce Type: new Abstract: We introduce $mathbf{SP-Rank}$, the first large-scale, publicly available dataset for benchmarking algorithms that leverage both first-order preferences and second-order predictions in ranking tasks. Each datapoint includes a personal vote (first-order signal) and a meta-prediction of how others will vote (second-order signal), allowing richer modeling than traditional datasets that capture only individual preferences. SP-Rank contains over 12,000 human-generated datapoints across three domains — geography, movies, and paintings, and spans nine elicitation formats with varying […]

Ver mais

arXiv:2503.18896v2 Announce Type: replace-cross Abstract: A statistical model is said to be calibrated if the resulting mean estimates perfectly match the true means of the underlying responses. Aiming for calibration is often not achievable in practice as one has to deal with finite samples of noisy observations. A weaker notion of calibration is auto-calibration. An auto-calibrated model satisfies that the expected value of the responses for a given mean estimate matches this estimate. Testing for autocalibration has only […]

Ver mais

Split learning (SL) enables collaborative training of large language models (LLMs) between resource-constrained edge devices and compute-rich servers by partitioning model computation across the network boundary. However, existing SL systems predominantly rely on first-order (FO) optimization, which requires clients to store intermediate quantities such as activations for backpropagation. This results in substantial memory overhead, largely negating benefits of model partitioning. In contrast, zeroth-order (ZO) optimization eliminates backpropagation and significantly reduces memory usage, but often suffers from slow convergence […]

Ver mais

arXiv:2511.19566v2 Announce Type: replace-cross Abstract: Open weight models, which are ubiquitous, rarely provide access to their training data or loss function. This makes modifying such models for tasks such as pruning or unlearning, which are constrained by this unavailability, an active area of research. Existing techniques typically require gradients or ground-truth labels, rendering them infeasible in settings with limited computational resources. In this work, we investigate the fundamental question of identifying components that are critical to the model’s […]

Ver mais