Experimental browser for the Atmosphere
We scaled training data attribution (TDA) methods ~1000x to find influential pretraining examples for thousands of queries in an 8B-parameter LLM over the entire 160B-token C4 corpus! medium.com/people-ai-re...
Dec 13, 2024, 6:57 PM
{ "uri": "at://did:plc:u4swdfntdfizkswnejceoi2y/app.bsky.feed.post/3ld7hmptt2s2a", "cid": "bafyreiclbc63wx6wfj2mnxra2deepcdlvsrvlsrow3rygn2miwgvdd6xqu", "value": { "text": "We scaled training data attribution (TDA) methods ~1000x to find influential pretraining examples for thousands of queries in an 8B-parameter LLM over the entire 160B-token C4 corpus!\nmedium.com/people-ai-re...", "$type": "app.bsky.feed.post", "embed": { "$type": "app.bsky.embed.images", "images": [ { "alt": "", "image": { "$type": "blob", "ref": { "$link": "bafkreics6ujprixgffniysm36dxj7utqtkwyqt4hyhnieep62rwj6hlve4" }, "mimeType": "image/jpeg", "size": 213260 }, "aspectRatio": { "width": 1180, "height": 574 } } ] }, "langs": [ "en" ], "facets": [ { "index": { "byteEnd": 210, "byteStart": 184 }, "features": [ { "uri": "https://medium.com/people-ai-research/scaling-training-data-attribution-f7d1eddd85da", "$type": "app.bsky.richtext.facet#link" } ] } ], "createdAt": "2024-12-13T18:57:01.182Z" } }