Dataless Knowledge Fusion by Merging Weights of Language Models · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2212.09849/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTDataless Knowledge Fusion by Merging Weights of Language Models — mentioned by sebastiennighthttps://news.ycombinator.com/item?id=41050687urn:hackerpapers:mention:2212.09849:41050687Tue, 23 Jul 2024 20:52:28 GMT<p>It is possible to merge two fine-tunes of models from the same family by... wait for it... averaging or combining their weights[0]. I am still amazed that we can do that. [0]: https://arxiv.org/abs/2212.09849</p><p><a href="https://hackerpaper.kaipfister.de/p/2212.09849/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CL