Quantized Reasoning Models Think They Need to Think Longer, but They Do Not · HackerPapers discussions

Quantized Reasoning Models Think They Need to Think Longer, but They Do Not — HN submission

Sun, 27 Sep 2026 19:12:57 GMT

Quantized Reasoning Models Think They Need to Think Longer, but They Do Not

13 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.

Quantized Reasoning Models Think They Need to Think Longer, but They Do Not — mentioned by kisjovan

Wed, 16 Sep 2026 14:26:56 GMT

I will TLDR you on our thought process, research, training and benchmarks. 1. When running our quantized Qwen 3.8 27B instances we were very annoyed by random reasoning loops (in the paper bellow refered to as "overthinking errors". These random loops were persistent throughout medium and low reasoning settings. 2. We found a paper by Meta that's supposed to target this phenomenon in PTQ, but when used straight out of the box got mixed results. https://arxiv.org/abs/2606.00206 3. We figured to try if it's a matter of the targeting the right keywords and tuning the parameters, so we used our…

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

Quantized Reasoning Models Think They Need to Think Longer, but They Do Not — mentioned by robertkarl

Wed, 17 Jun 2026 14:11:35 GMT

https://arxiv.org/abs/2606.00206 In this paper they nerf an LLMs ability to emit waffling thinking tokens like "wait", "but", "alternatively", and the models (they're old, small models in the paper) terminate reasoning faster and perform better. I bet Anthropic is tuning this on their backend.

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