Yu Li

Lumos-Nexus is a training-efficient framework that enhances video generation quality by progressively bridging generation from a lightweight model to a high-fidelity generator in a shared latent space, without sacrificing reasoning capabilities.

The paper introduces PIGMENT, a physics-informed foundation model that enables reliable quantitative mapping of brain microstructure from extremely sparse or challenging diffusion MRI scans.

The paper introduces SURE, a unified framework designed to standardize and improve the comparability and reproducibility of evaluations for advanced speech understanding models.

The paper introduces Lookahead Group Reward (&) to combat Supervision Fidelity Decay (SFD) in on-policy distillation, significantly improving student model performance on long reasoning tasks.

EvoDefense introduces an experience-guided, co-evolving black-box defense mechanism that significantly improves LLM robustness against unseen and diverse attacks without requiring model retraining.

The paper introduces Atomic Decomposition and Recombination (ADR), a novel framework that generates genuinely novel and challenging verifiable code tasks, significantly improving the scalability of Reinforcement Learning with Verifiable Rewards (RLVR) for LLMs.

MOSAIC introduces a structured agentic framework that treats automated data science as a staged, context-grounded model selection problem, improving performance and traceability over traditional AutoML and unconstrained LLM agents.

The paper introduces TukaBench, a culturally grounded jailbreak benchmark for seven African languages, demonstrating that prompting in African languages, especially with cultural adaptation, significantly reduces LLM refusal rates compared to English.

The paper introduces Expected Value Alignment (EVA), a novel reward modeling procedure that allows continuous scoring of intermediate reasoning steps in formal mathematics verification while maintaining the discrete, textual output format of generative models.

Soft-NBCE introduces soft entropy-weighted chunk fusion to overcome the semantic fragmentation caused by hard chunk selection in long-context LLMs, significantly improving performance on multi-hop benchmarks.

PolySpeech-100 introduces a massive, multi-lingual benchmark covering 110 linguistic variants to rigorously test Speech-LLMs, demonstrating that open-source models struggle with low-resource languages and that direct audio processing is superior to cascaded ASR+LLM systems.

The paper introduces Multi-temporal Referring Segmentation (MTRS), a new task requiring models to segment language-described temporal changes, and proposes MTRefSeg-R1, a specialized framework that achieves superior performance on the newly created MTRefSeg-21K benchmark.

The paper introduces PMC-InterCPT, a refined biomedical interleaved corpus that enhances multimodal continued pretraining by integrating figure-referencing body text alongside captions, leading to improved medical and general multimodal model performance.

TROPHIES introduces a unified framework to jointly reconstruct dynamic humans, static scenes, and camera poses from multi-view videos, achieving globally consistent and physically plausible 4D reconstructions.

The paper introduces InsightVQA, a large-scale benchmark dataset designed for hierarchical visual question answering that assesses complex emotion understanding and cognitive reasoning beyond simple emotion recognition.

The paper introduces CEON, a Circular Economy Ontology Network, designed to improve semantic interoperability and knowledge representation across diverse industry sectors throughout the product life cycle.

S-SPPO introduces a dual-space semantic calibration framework to stabilize Self-Play Preference Optimization (SPPO), preventing policy degeneration when preference oracles assign overly confident wins to semantically similar responses.

The paper introduces Bastet, a novel, high-quality, expert-labeled dataset designed to overcome limitations in existing resources for detecting complex smart contract vulnerabilities in DeFi.

This paper investigates the downlink performance of CoMS-NOMA networks from a system-level perspective.

This paper proposes Tail-Aware Adaptive-k (TAA-k), a training-free framework for adaptive context selection in retrieval-augmented generation systems using Extreme Value Theory.