MLINFO COLLECTION
Looped transformers reuse a shared block of layers to increase computational depth without increasing the number of trainable parameters. They are trained with backpropagation through time (BPTT).
Language models can make personalized recommendations from rich item content, user history, and feedback. However, evaluating many candidates with full-text inputs incurs substantial online…
In a hierarchical multi-agent system, a Leader assigns tasks and subagents carry them out. Jointly training the Leader and the subagents is difficult because a final answer gives only indirect…
Mixture-of-experts (MoE) models often exceed a single accelerator's memory capacity. Offloading experts to host memory creates a severe PCIe I/O bottleneck because data-dependent routing keeps…
In-context reinforcement learning (ICRL) exploits the impressive in-context learning ability of large language models to zero-shot generalize to new, unseen tasks at inference time without any…
Clinical registries and wearable sensors support early health-risk assessment, yet models trained on these records may leak sensitive information when released. Differential privacy (DP) bounds this…
Conformal prediction (CP) for language models is typically evaluated under a matched protocol, in which calibration and deployment use the same post-training inference configuration, such as…
In contact-rich manipulation, contact conditions can change before a visual policy updates its action plan. Recomputing the plan through the visual policy adds latency, while tactile feedback alone…
RLVR improves language-model reasoning, but commonly relies on independent stochastic rollouts for exploration, which can repeatedly sample similar failed strategies. Prior work has seen success with…
In human-robot interaction (HRI) tasks, humans may interact with the robot in either an active guiding or passive following state, resulting in distinct control objectives for each interaction mode.…
Reinforcement learning (RL) enables adaptive and autonomous decision-making in embedded, edge, and heterogeneous computing systems. Field-programmable gate arrays (FPGAs) and integrated circuits…
Most high-performing end-to-end autonomous driving agents (ADAs) are developed for the CARLA Leaderboard, tightly coupling them to CARLA's APIs, sensor layouts, coordinate systems, and internal…
Video generation models can now depict remarkably realistic motion. Yet it is unclear how to use this motion beyond pixels: to animate a specific 3D asset and replay the animation in its 3D scene.
Motivated by the human ability to integrate experience across language, perception, and action, we introduce UniWALM, a unified world-action-language model that learns video, action, and language in…
Planning under uncertain control execution noise requires reliable feedback policies across diverse environments. We introduce Planning Under Noise (PUN), a neural operator that learns noise-aware…
Robots exploring unknown indoor environments must decide where to move while much of the environment remains unseen. Predicting the occupancy of unseen space beyond the current sensor range can…
Envy-free cake cutting asks how to divide a resource so that no agent prefers another agent's share. Finding such a division with few queries about preferences has been a longstanding challenge.
Tracking large human-motion datasets has become the dominant route to general, expressive humanoid whole-body control, but scaling across many motions is expensive. Current pipelines use model-free…
Object-goal navigation (ObjectNav) requires an embodied agent to find an instance of a specified category in an unseen environment and stop near it. The agent typically relies on an object detector…
A neural model is the set of functions that can be expressed by a given neural network architecture. Neural models contain natural submodels, such as those obtained by setting some neurons to be…
Navigation Among Movable Obstacles (NAMO) requires a robot to rearrange objects that block access to its goal. For a pushing robot, progress may depend on a setup push that enables a later opening…
In reinforcement learning, one presumes the action set to be known ahead of time, and, thus, one must know if any state-action pairs are not permissible ahead of computations. In some applications,…
Generating graph signals is an important task in machine learning, and incorporating the underlying graph into a generative model enables structurally-informed distribution modeling. However, said…
Prior-data fitted networks (PFNs) have recently gained popularity as an end-to-end approach to amortized Bayesian inference. Yet, their statistical behavior for semi-parametric targets, such as means…