Stealing Reasoning Traces from Proprietary LLM APIs · HackerPapers discussions

Stealing Reasoning Traces from Proprietary LLM APIs — mentioned by irthomasthomas

Wed, 28 Oct 2026 10:53:33 GMT

I noticed last night that openai had fallen way down the index, and I allowed myself to think for a minute that maybe AA aren't as biased as I thought. Then a few hours later they push this fix which sees openai rise again. This update is even worse than usual because it now includes two private sources, data which we can't reproduce if we wanted to do. Also note that Anthropic is the only AI company actually caught red handed cheating at evals. Teaching their models to memorize test answers and lie about it, using their hidden reasoning tokens to hide the deception [0]. Yet we see no comment…

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

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Tue, 08 Sep 2026 14:41:49 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Thu, 20 Aug 2026 02:12:17 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Wed, 19 Aug 2026 17:19:22 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

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

Stealing Reasoning Traces from Proprietary LLM APIs — mentioned by johnnyApplePRNG

Mon, 17 Aug 2026 02:58:38 GMT

According to the paper "Stealing reasoning traces from proprietary llms" [0] all frontier models overthink. Thinking is good. You just don't see it in proprietary harnesses because it's literally cryptographically hidden from you. [0] https://arxiv.org/pdf/2608.09867

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

Stealing Reasoning Traces from Proprietary LLM APIs — mentioned by rikima_

Fri, 14 Aug 2026 23:36:15 GMT

This might be helpful https://arxiv.org/abs/2608.09867

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

Stealing Reasoning Traces from Proprietary LLM APIs — mentioned by xmcqdpt2

Fri, 14 Aug 2026 11:59:00 GMT

pdf https://arxiv.org/pdf/2608.09867 for some reason I couldn't find any way to download it from that website.

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

Stealing Reasoning Traces from Proprietary LLM APIs — mentioned by irthomasthomas

Fri, 14 Aug 2026 09:37:42 GMT

Have you seen the news about decrypting the hidden COT in U.S. models? [0] The decoded logs revealed instances where Claude memorized answers to test questions beforehand while making its final output look like it had derived the answer step-by-step—hiding the memorization from the user. 0: https://www.alphaxiv.org/abs/2608.09867?hl=en-GB

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

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Thu, 13 Aug 2026 07:41:03 GMT

Stealing Reasoning Traces from Proprietary LLM APIs [pdf]

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Wed, 12 Aug 2026 23:09:42 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

5 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Wed, 12 Aug 2026 14:17:03 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

6 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Wed, 12 Aug 2026 01:56:49 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Tue, 11 Aug 2026 15:23:14 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

9 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Stealing Reasoning Traces from Proprietary LLM APIs — HN submission

Tue, 11 Aug 2026 07:07:06 GMT

Stealing Reasoning Traces from Proprietary LLM APIs

5 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.