3 papers
cs.LG2026
Trust but Verify: Designing Region-Inhomogeneous Forward Dynamics for Discrete Diffusion
Gunmay Jhingran
Continuous diffusion treats the forward process as a design space; discrete language diffusion often uses absorbing corruption, leaving context fully trusted throughout generation.…
cs.CL2026
Rules Amortize, Pairings Don't: Linguistic Structure Determines What Latent Task Representations Can Replace In-Context Learning
Gunmay Jhingran
In-context learning (ICL) can be amortized into latent objects (task vectors, function vectors, context vectors) that recover few-shot behavior at zero-shot inference cost, but rec…
cs.LG2025
Deterministic Continuous Replacement: Fast and Stable Module Replacement in Pretrained Transformers
Rowan Bradbury, Aniket Srinivasan Ashok, Sai Ram Kasanagottu +2
Replacing modules in pretrained models, especially swapping quadratic self-attention for efficient attention alternatives, poses a hard optimization problem: cold-start reinitializ…