Akash Srivastava

Mitigating Premature Exploitation in Particle-based Monte Carlo for Inference-Time Scaling

Sculpting Subspaces: Constrained Full Fine-Tuning in LLMs for Continual Learning

Inference-Time Scaling of Diffusion Language Models with Particle Gibbs Sampling

CosyCPT: Coreness-Aware Synthetic Continued Pretraining

Rollout Roulette: A Probabilistic Inference Approach to Inference-Time Scaling of LLMs using Particle-Based Monte Carlo Methods

Activation-Informed Merging of Large Language Models

SQuat: Subspace-orthogonal KV Cache Quantization

Unveiling the Secret Recipe: A Guide For Supervised Fine-Tuning Small LLMs

Discrete Inversion: A Controllable Latent Space for Multinomial Diffusion and Masked Generative Models

Preference Data Annotation with Guided Density Ratios