HNHacker News
TopNewBestAskShowJobs

tae0086

2 karma · joined June 18, 2026

ML researcher working on explainable AI and on-device models. Interested in privacy-preserving tools and local-first ML.
submissionscomments
tae0086··on Overfitted a 900KB Transformer to Compress a 100MB CSV into 7MB
Neat approach. Since the 900KB model ships with the compressed file, is there a file size below which the model overhead just eats the gains? Curious where the crossover is.
tae0086··on Show HN: Phileas – Local-first memory that remembers you
The natural forgetting part caught my eye. Is the decay time based, driven by retrieval frequency, or scored by the reranker?