「打好关系、要做汇报、出同款」:AI 检索初创 SID 指控向量数据库 Chroma 抄袭其研究成果

BlockBeatNews

According to 1M AI News monitoring, SID.ai CEO Max Rumpf published a long post on X in which he openly accused the open-source vector database Chroma’s newly released Context-1 model of heavily borrowing SID’s research成果 published by SID in December last year, without providing any citations or acknowledgments.

Rumpf shared email exchanges with Chroma CEO Jeff Huber as evidence. In October 2025, Huber proactively asked Rumpf what model he was training. Rumpf replied that he was working on an “agentic retrieval model—similar to Cognition’s SWE-grep but for general-purpose retrieval—which is already stronger than Sonnet 4.5 and Gemini 2.5 Pro.” After the SID-1 technical report was released in December 2025, Rumpf again shared a link with Huber, and Huber replied, “Congratulations.” Both companies are YC alumni, and their offices are adjacent.

Both SID-1 and Context-1 are agentic retrieval models trained with reinforcement learning. Both are positioned as retrieval sub-agents for frontier reasoning models. Both use synthetic data for training. Both claim to achieve the Pareto frontier in cost and latency. The specific similarities Rumpf lists include: Figure 1 using the same speed/cost dual-view switching, 4-way parallel inference combined with RRF (reciprocal rank fusion) to aggregate results, and the overall framework of the charts, datasets, and methodology.

Context-1’s technical report cites related-work research in the same field such as WebExplorer, SWE-grep, and Search-R1, but the full text does not mention SID-1, and benchmark evaluations do not include SID-1 in the comparisons. Rumpf says that Chroma, “knowing there is another model,” still claims “Pareto optimal,” and points out that although Context-1 has open-sourced the weights, the inference framework required to run it has not been released, preventing SID from benchmarking it.

Rumpf said this approach “completely destroys the motivation we (and others) had to share in-depth in technical reports,” and called it “a regrettable bad research practice in academia that is spreading to startups.” As of the time of this writing, Chroma had not publicly responded.

Disclaimer: The information on this page may come from third parties and does not represent the views or opinions of Gate. The content displayed on this page is for reference only and does not constitute any financial, investment, or legal advice. Gate does not guarantee the accuracy or completeness of the information and shall not be liable for any losses arising from the use of this information. Virtual asset investments carry high risks and are subject to significant price volatility. You may lose all of your invested principal. Please fully understand the relevant risks and make prudent decisions based on your own financial situation and risk tolerance. For details, please refer to Disclaimer.
Comment
0/400
No comments