Experimental browser for the Atmosphere
The first challenge was to create a dataset large enough to train a language model. Our researchers were committed to making it fully open, consulting with legal and ethics experts early in the process to create dataset design principles they would follow.
May 6, 2025, 8:55 PM
{ "uri": "at://did:plc:i4kytxgsu3yfsrt2ml3o7tgq/app.bsky.feed.post/3lojrfdvhc22c", "cid": "bafyreicxozwbtkc4av56ew5y2nstzxx4h62wnwadeaxtj5xjmw4kgolrzq", "value": { "text": "The first challenge was to create a dataset large enough to train a language model. Our researchers were committed to making it fully open, consulting with legal and ethics experts early in the process to create dataset design principles they would follow.", "$type": "app.bsky.feed.post", "langs": [ "en" ], "reply": { "root": { "cid": "bafyreifmvesjgu7dbxpy7x6q72n6wmwkgvvj6ae6awysunnjhi6vbdpk3y", "uri": "at://did:plc:i4kytxgsu3yfsrt2ml3o7tgq/app.bsky.feed.post/3lojrfdvahc2c" }, "parent": { "cid": "bafyreidpysjgvmvve2uusvr5wuj7ksu4je4yf2gj6j2nnv2nuevftausju", "uri": "at://did:plc:i4kytxgsu3yfsrt2ml3o7tgq/app.bsky.feed.post/3lojrfdvgcs2c" } }, "createdAt": "2025-05-06T20:55:36.475Z" } }