6 papers
How Far Does Second-Order Linear Response Predict Quantization Mismatch? Pre-Registered Measurements on Five Language Models
Dezhen Lu, Zixiang Ni
Serving systems choose a weight format for each model and deployment. A low-cost screen predicts the resulting KL divergence from the unquantized model's Fisher information using $…
What Sets the Serving Speed of Looped Language Models? Measurements and Optimizations Graded by Exactness
Dezhen Lu, Runguo Li, Zixiang Ni +5
Looped language models apply one block of layers many times per token, so a serving engine pays every per-layer overhead once per loop; we measure serving for Ouro and Huginn-0125…
Where KV-Cache Error Lives in Looped Language Models: A Diagnostic Study of Ouro and Huginn
Dezhen Lu, Runguo Li, Zixiang Ni +5
A looped language model applies one block of layers 4 times per token and keeps a key/value (KV) plane for each loop. We study where storage error changes the next-token distributi…
ReTaCo: Residual-Target Control for On-Policy Distillation
Zixiang Ni, Zhuo Hu, Renjie Cao +8
On-policy distillation (OPD) trains a student on its own generated prefixes with token-level teacher feedback, but transmitting or storing the teacher's full-vocabulary distributio…
RADIANCE: Relative Adaptive Denoising with IP-Adapter for Novel Concept Enhancement
Zi-Xiang Ni, Bo-Lun Huang, Teng-Fang Hsiao +2
Text-to-image (T2I) diffusion models have achieved striking progress but still struggle to synthesize rare concepts involving unusual attribute-object pairings, often resulting in…
Score Replacement with Bounded Deviation for Rare Prompt Generation
Bo-Kai Ruan, Zi-Xiang Ni, Bo-Lun Huang +2
Diffusion models achieve impressive performance in high-fidelity image generation but often struggle with rare concepts that appear infrequently in the training distribution. Prior…