Co-evolves candidate solutions and web-search queries to help LLM-driven evolutionary discovery, using a retrieval gate plus bilevel inner/outer loops that refine queries, rank documents by predicted solution value, and generate evaluated candidates.
Predicts compact 'prospective tokens' that summarize upcoming information needs and uses them to select a small set of past frames for conditioning long-horizon video generation, improving long-range consistency, visual quality, and action alignment while remaining plug-and-play across diverse generators.
Calibrates multi-reward reinforcement learning by adaptively upweighting infrequently active rewards per rollout batch, so sparse objectives provide stronger signals when they matter. Proposes an inverse-square-root density correction and shows faster learning on tool-calling and math-reasoning tasks.
Generates stereo audio whose perceived locations evolve consistently with visible source motion in joint video–audio generation. Introduces StereoBind (motion-conditioned stereo synthesis), a large-scale training set StereoWorld-29K, and an evaluation suite StereoWorldBench for dynamic spatial correspondence.
Uses a multimodal model's own critiques as privileged context and applies on-policy self-distillation over diffusion sampling trajectories to internalize corrective guidance, improving text-to-image generation without an external teacher; shows measurable gains on GenEval and GenEval2.
Analyzes how proposer–solver loops in self-evolving search agents can develop shared errors (co-cheating) that inflate internal rewards; introduces Multi-Sample Verification and CrossFit (cross-fitted scoring with partitioned sources) to reduce false agreement and improve downstream search performance.
Assesses how AI reviewers react to content-preserving rewrites and introduces RobustReview, a controlled benchmark plus SciCore, a dual-branch reviewer that combines full-manuscript judgment with a structured 'science core' summary to reduce rhetorical sensitivity.
Demonstrates that additive weighted aggregation in rubric-based reinforcement learning enables reward hacking—models improve rubric coverage but produce worse answers—and proposes ProRubric, an offline regrouping that aggregates checklist criteria into conjunctive protocol-level dimensions to raise appropriateness without losing coverage.
Constructs and continually maintains explicit belief states for long-horizon LLM agents, combining a structured world estimate with unresolved epistemic and achievement gaps. Adds consistency validation, Belief Trapping detection, and tailored recovery to improve execution and diagnosis benchmarks.
Combines joint distribution distillation from a video teacher with marginal (frame-level) distillation from an image teacher to improve few-step video generation. Introduces LatentBridge to align incompatible latents and Latent Variation Sampling to distribute frame supervision, boosting per-frame visual quality and semantic alignment while largely preserving motion.
Analyzes how LLM agents prefer items from particular sources during end-to-end search and how these preferences affect selections across shopping, accommodation, and scholarly domains. Shows source labels can override item quality and evaluates mitigation strategies such as hiding sources, relabeling, supplying missing information, and counter-prompts.
Introduces a token-adaptive latent recurrence for diffusion language models that allocates computation per token during denoising. Key features: iterative latent refinement, discrete-feedback commits, and learned token-wise schedules to improve parallel decoding quality-efficiency trade-offs.