Search papers, labs, and topics across Lattice.
Chinese Information Processing Laboratory, Institute of Software, Chinese Academy of Sciences
14
0
14
25
Even the most advanced autonomous agents struggle with maintaining document consistency, revealing critical failure modes that could hinder their effectiveness in real-world applications.
ShortOPD boosts the generative performance of pruned LLMs by nearly 9 times while cutting training time by over 75%.
Transforming historical sequences into a powerful resource, PraMem significantly improves long-horizon behavior prediction beyond existing methods.
ReasoningLens turns the opaque reasoning of large models into clear, actionable insights, enabling researchers to pinpoint errors and optimize performance like never before.
Recursive composition of verifiable environments can boost reasoning performance in RL by up to 3.1 points while using only a fraction of the original environments.
Current AI agents falter in autonomous development, revealing critical gaps in robustness and alignment as they struggle against human-engineered solutions.
ADR transforms the landscape of code task generation, enabling LLMs to tackle genuinely novel and challenging coding problems that enhance their performance.
Forget scraping – this work shows you can generate high-quality, executable terminal environments from scratch to train language agents that outperform models trained on scraped data.
MLLMs can't grasp metaphors in videos, revealing a surprising gap in their high-order cognitive abilities compared to humans.
LLMs trained with ScaleBox, a new high-fidelity code verification system, substantially outperform those trained with heuristic matching, suggesting current RLHF methods are bottlenecked by verification quality.
Multilingual RAG systems are systematically suppressing "answer-critical" documents in non-English languages, crippling their ability to leverage global knowledge.
Forget text-dominance: Today's Omni-modal LLMs surprisingly favor visual inputs, creating new challenges for cross-modal reasoning.
LLMs exhibit a "Utopian bias" when simulating human behavior, converging towards an unrealistic "positive average person" and failing to capture individual differences and long-tail behaviors.
By grounding reflection in the visual artifacts of presentation slides, DeepPresenter enables agents to iteratively refine presentations in a way that internal reasoning traces alone cannot.
