Joshua Kimball — accepted AI-conference papers

2026

DPAR: Dynamic Patchification for Efficient Autoregressive Visual Generation

Divyansh Srivastava, Akshay Mehra, Pranav Maneriker, Debopam Sanyal, Vishnu Raj, Vijay Kamarshi +2

CVPR 2026

Decoder-only autoregressive image generation typically relies on fixed-length tokenization schemes whose token counts grow quadratically with resolution, substantially increasing the computational and memory demands of attention. We present DPAR, a novel decoder-only autoregressive model that dynami

Cited by 0 Source ↗code ↗Scholar ↗

2025

Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning Attack

Tiansheng Huang, Gautam Bhattacharya, Pratik Joshi, Joshua Kimball, Ling Liu

ICML 2025poster

Safety aligned Large Language Models (LLMs) are vulnerable to harmful fine-tuning attacks -- a few harmful data mixed in the fine-tuning dataset can break the LLMs's safety alignment. While several defenses have been proposed, our evaluation shows that existing defenses fail \textit{when some specif…

Cited by 0 Source ↗PDF ↗Scholar ↗