memory
Agentic Memory
Continual learning of AI agents — in-context learning, continual fine-tuning, and unlearning.
directions
world model
World Model
In-context world models, adaptation to post-training task worlds, and adapting agents in evolving envs.
latency
Low-Latency AI
Efficient attention architectures, KV-cache compression, latent segmentation, and recurrent transformers.
safety
AI Safety
Synthetic data training, risks of multi-agent interaction, post-training guardrails, and AI behavioral study.
recent papers
- 012026memoryin submission · NeurIPS 2026
SHRED: Document Unlearning via Self-Distillation and Entropy Demotion
A document-level unlearning method that combines self-distillation on retain data with entropy demotion on the forget set. Removes targeted knowledge from LLMs without catastrophic damage to unrelated capabilities.
view - 022026safetyin submission · EMNLP 2026
Expert Personas Improve LLM Alignment but Damage Accuracy: Bootstrapping Intent-Based Persona Routing with PRISM
Persona effectiveness is task-type dependent: expert prompts consistently improve alignment-dependent tasks (safety, preference) but reliably damage pretraining-dependent knowledge retrieval. PRISM teaches models when to invoke a persona via intent-based self-modeling, preserving accuracy while keeping alignment gains.
view - 032026latencyin progress
AttendTwice: Long-Context Inference via Dynamic Token-Level KV-Cache Selection
A two-pass attention scheme that dynamically selects which KV-cache tokens to attend to per query, enabling long-context inference at a fraction of the standard memory footprint.
- 042025safetyACM ICMI 2025
Multimodal Synthetic Data Finetuning and Model Collapse
Studies how vision-language models degrade when fine-tuned on AI-generated multimodal data. Characterizes the collapse dynamics specific to the multimodal regime and proposes mitigation strategies that preserve diversity across modalities.
view - 052024latencypreprint
Lateralization MLP: A Simple Brain-inspired Architecture for Diffusion
A brain-inspired MLP architecture with hemispheric lateralization applied to diffusion models. Shows competitive sample quality at reduced parameter count, suggesting structured asymmetry as an inductive bias for generative modeling.
view - 062024latencypreprint
Static Key Attention in Vision
A more efficient attention variant for vision transformers that pre-computes a static key projection, reducing per-token compute while maintaining downstream task performance.
view
academic service
- Reviewer · NeurIPS 2024–2026
- Reviewer · ICLR 2024–2025
- Reviewer · ICML 2024–2025
- TA · DSCI 552 (USC)