The paper introduces the Latent Interconnect Protocol (LIP), a framework that allows large language models (LLMs) to directly exchange rich, high‑dimensional representations with specialist models (e.g., protein and small‑molecule experts) via learned connectors. By training only these connectors, the approach enables zero‑shot transfer to cross‑domain tasks and compositional reasoning without joint training of the LLMs or specialists. Experiments with Gemma 3 and Kimi K2.6 show improved task performance and compositionality, especially in zero‑shot settings, while highlighting the need for consistent specialist interfaces and LLM training to handle such inputs.
arXiv:2608. 12090v1 Announce Type: new Abstract: Protein language models (PLMs) have transferred the latest advances from natural language processing to computational biology.
ReVisIT is a train‑free framework that turns retrieved image‑label pairs into units of visual thought, combining structured class definitions, multimodal retrieval, and alternating user/assistant injection before joint decoding. On several benchmarks—including Fast Open MiniImageNet, Bongard‑OpenWorld, and the newly released MAAC‑Bench—ReVisIT achieves performance comparable to or surpassing large, trained models while using far fewer parameters. The approach demonstrates that high‑quality retrieval and a simple turns layer can provide a universal performance boost across diverse multimodal tasks.
arXiv:2601. 13591v2 Announce Type: replace Abstract: Recent LLM-based data agents aim to automate data science tasks ranging from data analysis to deep learning.
Protein language models (PLMs) have transferred the latest advances from natural language processing to computational biology. These models, trained on large corpora of protein sequence data, are widely used to translate amino acid sequences into latent-space embeddings, ready for use in diverse downstream tasks (DTs).
SkillMem introduces a compact external latent memory for reusable skills in latent multi‑agent systems, enabling role‑conditioned cross‑attention to guide agents without storing large per‑skill parameters. Trained via behavioral distillation, SkillMem matches the effect of textual skills while keeping the backbone frozen. Across six benchmarks and Qwen3 models from 4B to 14B, it delivers strong gains for smaller backbones, reduces skill storage by over 38×, and achieves up to 4.23× inference speedup.
The paper introduces UniSandbox, a decoupled evaluation framework with controlled synthetic datasets, to study whether understanding informs generation in Unified Multimodal Models. Results show a notable understanding‑generation gap, especially in reasoning generation and knowledge transfer. Explicit Chain‑of‑Thought (CoT) in the understanding module bridges this gap, and self‑training can internalize CoT for implicit reasoning during generation; query‑based architectures also exhibit latent CoT‑like properties that aid knowledge transfer.
arXiv:2609.00474v1 Announce Type: cross Abstract: LLMs are increasingly deployed as orchestrators that coordinate specialized subagents to solve complex tasks through natural language. However, in ma...
arXiv:2605.10120v2 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) show remarkable potential for scientific reasoning, yet their performance in specialized domains suc...
arXiv:2608. 13505v1 Announce Type: new Abstract: Scientific discovery increasingly requires AI systems that can reason over scientific evidence of heterogeneous modalities, interact with scientific tools and environments, and sustain progress across long task horizons.
arXiv:2511. 20639v3 Announce Type: replace-cross Abstract: Multi-agent systems (MAS) extend large language models (LLMs) from independent single-model reasoning to coordinative system-level intelligence.
ToM2Vec introduces continuous theory-of-mind tokens that enable large language model agents to reason about multiple possible partner states in multi‑agent settings. By training these tokens to abstract partner thoughts, the model generates latent representations that condition agents’ thoughts and actions, improving performance across collaborative, competitive, and mixed‑motive environments. The approach demonstrates that latent reasoning outperforms text‑based reasoning and generalizes to unseen partner models, with tokens encoding partner beliefs, desires, and next actions.
Open multimodal reasoning models have benefited from large-scale reasoning supervision, yet reliable post-training remains challenging due to uneven data quality, inefficient supervision construction,...