There are many reasons one may want to edit a model’s chain of thought (CoT)—e.g., to prefill it with reasoning from a stronger model or to remove steps that may yield unsafe outputs. The success of these edits plausibly depends on a model’s inability to notice them, as the model may alter its behavior if it suspects tampering. In this work, we study whether recent reasoning models are able to detect such edits on their CoTs under a variety of conditions: during reasoning and after it; when alerted to the possibility of tampering and when not alerted; and when prefilled both with their own CoTs and with those of other models. Broadly, we find that (i) models exhibit only very modest detection accuracy; (ii) models struggle to identify *where* their CoT was modified; and (iii) models are about as good at detecting changes to their own CoTs as to those of other models.

A reliable resume-job matching system helps a company recommend suitable candidates from a pool of resumes and helps a job seeker find relevant jobs from a list of job posts. However, since job seekers apply only to a few jobs, interaction labels in resume-job datasets are sparse. We introduce ConFit v2, an improvement over ConFit to tackle this sparsity problem. We propose two techniques to enhance the encoder’s contrastive training process: augmenting job data with hypothetical reference resume generated by a large language model; and creating high-quality hard negatives from unlabeled resume/job pairs using a novel hard-negative mining strategy. This method also simplifies the representation space of the encoder. We evaluate ConFit v2 on two real-world datasets and demonstrate that it outperforms ConFit and prior methods (including BM25 and OpenAI text-embedding-003), achieving an average absolute improvement of 13.8% in recall and 17.5% in nDCG across job-ranking and resume-ranking tasks.