arXiv AI

GRIP: Grounded Reasoning via Information-Restricted Premises

GRIP (Grounded Reasoning via Information-Restricted Premises) addresses the query dominance problem in retrieval-augmented generation by enforcing a capacity asymmetry: the decoder retains full access to the query while retrieved evidence is funneled through a severe stochastic bottleneck. This design forces the evidence channel to encode only residual information not present in the query. On five reasoning benchmarks, GRIP surpasses strong iterative baselines, reduces query–latent mutual information by about 30×, cuts hallucination by 73%, and its bottleneck outputs occupy subspaces less aligned with the query than baseline representations.

arXiv AI

6d ago

Beyond Post-Hoc Rewards: Privileged Reasoning Exploration for Multimodal Embeddings

The paper introduces PILOT, a framework that treats reasoning as a latent variable to enhance multimodal embeddings. By modeling prior and posterior reasoning modes with a shared large language model, PILOT uses relational supervision during training to guide exploration of retrieval‑relevant reasoning traces. On the MMEB‑V2 benchmark, PILOT outperforms baselines across all task groups, notably improving visual‑document retrieval scores by over six points.

By Tianyu Yang, Yijun Tian, Lars Benedikt Kaesberg, Jan Philip Wahle, Bela Gipp, Terry Ruas, Daniel Kurzawe

llms rag multimodal benchmarks

arXiv AI

Aug 17

Retrieval Grounding Latent Reasoning for Dense Retrieval

arXiv:2608. 14107v1 Announce Type: new Abstract: Reasoning-intensive retrieval requires text representations to capture not only semantic similarity, but also the reasoning needed to determine relevance under a given retrieval instruction.

By Gang Zhou, Xiongxi Yu, Hu Tian, Yang Wei, Lu Pan, Ke Zeng, Shibiao Xu, Xiaolong Zheng

rag efficiency benchmarks

arXiv AI

Jun 6

Closing the Loop on Latent Reasoning via Test-Time Reconstruction

arXiv:2606. 06252v1 Announce Type: new Abstract: Recent work moves intermediate reasoning from natural-language traces into latent or cache-level representations to reduce token overhead and avoid a discrete communication bottleneck.

By Xiaopeng Yuan, Haibo Jin, Ye Yu, Peng Kuang, Lijun Yu, Yushun Dong, Haohan Wang

benchmarks

arXiv AI

Sep 16

State of Thought Enables Endogenous Reasoning

arXiv:2609.16055v1 Announce Type: cross Abstract: Test-time compute has emerged as a major approach to improving the capabilities of Large Language Models (LLMs). However, existing test-time reasonin...

By Zhiren Gong, Yikun Hou, Zihao Zeng, Ming Xiao, Chau Yuen, Wei Yang Bryan Lim

llms rag multimodal

arXiv AI

Sep 17

Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning

The paper introduces LGM, a neuro‑symbolic framework that disentangles long‑term memory by mapping historical interactions into a continuous latent graph. Instead of static memory graphs, LGM uses a sparse autoencoder to create query‑aware latent nodes and edges, then applies a graph encoder conditioned on the query to perform non‑linear message passing. Experiments on long‑term personalization benchmarks show that LGM outperforms existing methods in capturing both explicit and implicit user preferences and generating personalized responses.

By Cai Ke, Xinghao Chen, Xiaoyu Shen, Keyu Chen, Siyu An, Junnan Dong, Ruifeng Xu, Ruizhi Qiao, Xing Sun

rag agents diffusion benchmarks

arXiv Computation and Language

Sep 10

Benchmarking Hybrid Deep Research Across Database Querying and Web Search

arXiv:2609.09410v1 Announce Type: new Abstract: While autonomous agents have made significant strides in "deep research" by iteratively navigating the open web to synthesize information, real-world p...

By Ruofan Wu, Peiran Xu, Xiaolong Li, Fan Shu, Soyoung Yoon, Yite Wang, Xiaodong Yu, Boyi Liu, Feng Yan, Debiao Li, Yuxiong He, Zhewei Yao

llms agents benchmarks data-engineering

arXiv AI

Jul 17

Enhancing Small Language Models Reasoning through Knowledge Graph Grounding

arXiv:2607. 14149v1 Announce Type: new Abstract: Although large language models (LLMs) have set benchmarks for zero-shot reasoning, their deployment remains cost-prohibitive and environmentally taxing.

By Dimitrios Kelesis, Konstantinos Bougiatiotis, Georgios Paliouras

llms agents benchmarks

arXiv Computation and Language

Oct 22

HybridResearcher: Bridging the Structured-Unstructured Boundary for Deep Research Agents

arXiv:2610.24899v2 Announce Type: replace-cross Abstract: While LLM agents are increasingly capable of gathering and synthesizing evidence from a single homogeneous source, such as structured tables...

By Soyoung Yoon, Peiran Xu, Han Wang, Ruofan Wu, Xiaodong Yu, Feng Yan, Mohit Bansal, Debiao Li, Seung-Won Hwang, Yuxiong He, Zhewei Yao

llms agents reinforcement-learning fine-tuning benchmarks data-engineering

arXiv AI

Oct 15

PANINI-DR: Controlled Deep Research through Generative Semantic Workspaces

arXiv:2610.15208v1 Announce Type: cross Abstract: Deep research requires connecting multiple steps of search, reasoning, and computation, with intermediate findings guiding what to investigate next....

By M. Yigit Turali, Shreyas Rajesh, M. Efe Lorasdagi, Chenda Duan, Kartik Sharma, Vwani Roychowdhury

agents benchmarks

arXiv Machine Learning

Sep 22

Assessing Adversarial Robustness of Latent Reasoning Models

arXiv:2609.22228v1 Announce Type: cross Abstract: Large language models increasingly rely on long chain-of-thought (CoT) trajectories for complex reasoning, but autoregressive generation brings subst...

By Shaolong Chen, Ang Li, Mingjie Li, Yisen Wang

llms multimodal benchmarks safety

arXiv AI

Sep 1

Beyond Dense States: Sparse Transcoders as Causally Testable Operators for LLM Latent Reasoning

arXiv:2602.01695v2 Announce Type: replace Abstract: Latent reasoning reduces the token-generation cost of chain-of-thought reasoning by replacing explicit intermediate tokens with continuous latent t...

By Yadong Wang, Haodong Chen, Yu Tian, Chuanxing Geng, Dong Liang, Xiang Chen

llms benchmarks

arXiv AI

Aug 11

Reason Wide, Not Deep: Amortizing the Reasoning Premium into Distilled Skills

arXiv:2608. 07885v1 Announce Type: new Abstract: Reasoning modes of language models outperform their non-reasoning counterparts on multi-step agentic tasks, but pay a 3-6x premium in output tokens on every episode -- much of it spent re-deriving procedures that are shared across episodes of the same domain.

By Agamdeep Singh, Srishti Gautam, Priyanshu Gupta, Nikita Mehrotra, Tanmay Bakshi, Sumit Gulwani

llms agents efficiency benchmarks

Related stories