…positional biases and attention sinks cause the model to allocate most of its attention to positionally privileged tokens rather than semantically relevant content. This training-time attention dilution (the starvation of content…
… We instantiate this on Formula 1 strategy explanations grounded in telemetry multilingual EN/ES/PT , where the oracle is derived from public timing data. …
… Existing benchmarks also suffer from data contamination, label noise, and degenerate binary relevance. …
… Second, it prevents hallucinated sources by treating retrieval as intent--paper relevance ranking rather than generation. …
…In this setting, a conversational AI assistant can access a user's personal camera roll and retrieve relevant photos to answer queries, ranging from simple factual questions (e.g., ``Name of the…
…We construct image-language-3D flow triplets from heterogeneous human and robot videos, and use these triplets as training-time supervision to shape an image-only encoder. Our key idea is to…
…It then optimizes a steering matrix to identify behavior-relevant expert circuits and, at inference time, applies steering masks to the routing gates to override expert selection. This enables targeted enhancement or…
… Given a budget of k frames, PEEK predicts per-frame relevance scores and returns the selected frames in temporal order, ready to be forwarded to a downstream Video-Language Model. …
…under strong angular intervention, a modest expansion of this capacity relieves competition between the steered concept and other context-relevant features, accounting for the observed stability gain. Paper: https://arxiv.org/abs…
…reasoning and then selects question-relevant frames adaptively, resulting in sub-quadratic attention cost with little to no accuracy loss. We arrange frames on a K{times}K grid and run lightweight…