p-less Sampling: A Robust Hyperparameter-Free Approach for LLM Decoding — mentioned by Der_Einzige

Tue, 21 Jul 2026 11:09:03 GMT

Switching to mlx-vlm is basically harmful since it (like vllm and sglang) have such garbage support for modern samplers. To be clear, I am one of the authors on the min_p paper, and if min_p is the best you have (when llamacpp supports the far superior top-n-sigma), than I have no reason to switch even if you are somehow faster. (https://arxiv.org/abs/2411.07641) And if you do care to support modern samplers, you can start with the following: 1. https://arxiv.org/abs/2509.23234 2. https://arxiv.org/abs/2509.02510 3. https://arxiv.org/abs/2604.11012

Paper and discussion history. A mention is not an endorsement.

p-less Sampling: A Robust Hyperparameter-Free Approach for LLM Decoding · HackerPapers discussions

p-less Sampling: A Robust Hyperparameter-Free Approach for LLM Decoding — mentioned by Der_Einzige