LLMs + Persona-Plug = Personalized LLMs (2025 · 31 citations)
Making LLMs adapt to individual users usually means either training a separate model per user, which is costly and fails when user data is scarce, or retrieving a few relevant past texts as demonstrations.
Selective Preference Optimization via Token-Level Reward Function Estimation (2025 · 8 citations)
Token-level preference alignment for LLMs usually trains on every token, which adds noise and wastes compute.
Tiny Budgets, Big Gains: Parameter Placement Strategy in Parameter Super-Efficient Fine-Tuning (2025 · 8 citations)
Parameter-efficient fine-tuning methods such as LoRA lower the cost of adapting large models. Yet models now reach hundreds of billions of parameters, so even rank-1 LoRA can be too heavy for mobile hardware or labs with little compute. The authors ask how a very small, fixed budget of trainable parameters should be…
Learning from Cognition: Enhancing RL Efficiency for LLM Reasoning via Hierarchical Metacognitive Decomposition and Refinement (2026)
Zero-RL training with verifiable rewards wastes much of its compute on weaker base LLMs. With a fixed prompt template, many problems yield no correct rollout, so accuracy-based filtering discards them. Cog-Rethinker tackles this by adding two cognition-inspired rollout stages on top of a DAPO-style pipeline. Problems…

Co-authors here

12 of 18 co-authors with a page, the most papers together first.

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

Papers of our library cite their works 35 times (of the 149,439 papers whose bibliographies we read); 14 citing papers with a page here →.

Updated 1 November 2026