…Our approach leverages a shared bidirectional transformer encoder to jointly represent text, entity type labels, and relation type labels, enabling zero-shot extraction of arbitrary entity and relation types specified at inference…
…A Globally Shared Expert Pool for Mixture-of-Experts Published on May 7 Submitted by Minbin Huang on May 8 CUHK Authors: , , , , , , , Abstract UniPool introduces a shared expert pool architecture for Mixture…
…View arXiv page View PDF Project page GitHub 1 Add to collection Community Excited to share our paper on On-Policy Distillation! No model linking this paper Cite arxiv.org/abs/2605…
…To quantify the security risks faced by LLMs under these threats, the research community has developed heuristic-based and optimization-based red-teaming methods. Optimization-based methods generally produce stronger attacks than…
…small trainable adapters as persistent local state on top of strong shared foundation models . In this framing, the base model provides shared competence while adapters carry instance-specific behavior such as preferences…
…View arXiv page View PDF Project page GitHub 40 Add to collection Community Excited to share EMO, a new MoE model pretrained so modular expert structure emerges directly from data, without human…
…View arXiv page View PDF GitHub 3 Add to collection Community the generalist embedding model that unifies tabular classification and retrieval in a single shared space. the cross-modal setup in tabembed…
…View arXiv page View PDF GitHub 0 Add to collection Community Excited to share CIPER (Cross-view Image-retrieval and Pose-estimation transformER)! Cross-view geo-localization—locating a ground image via…
…Recent studies show that the shared representation space of pretrained multimodal contrastive models can serve as a bridge, enabling models to perform multimodal training with unimodal data. However, the key premise of…
…VideoMLA replaces per-head keys and values with a shared low-rank content latent and a shared decoupled 3D-RoPE positional key, reducing per-token KV memory by 92.7% at every…