Langsung ke konten

ByteAugur

@byteaugur.bsky.social did:plc:n6yyiol7imf5r7e75gpu6eld

Fortune teller for the feed. I read leaks, patch notes, model drops and earnings to spot what’s next in AI, tech & gaming. linktr.ee/byteaugur

Bergabung dengan Bluesky pada

Pengikut
60
Mengikuti
300
Postingan
344
Hari di papan peringkat
0

Ringkasan

ByteAugur (@byteaugur.bsky.social) adalah pengguna Bluesky yang bergabung pada 21 Oktober 2026. Memiliki 60 pengikut, mengikuti 300 akun, dan telah memublikasikan 344 postingan.

Kemunculan di papan peringkat

Hari-hari terbaik saat postingan akun ini masuk daftar teratas harian Debya.

Akun ini belum pernah muncul di papan peringkat harian.

Postingan yang baru disukai

Suka publik dari jaringan Bluesky.

  • Sergey Baus@sergeybaus.com

    I made Telesec, a Hermes Agent plugin for Telegram Business: voice and video notes become text in the same chat. My short voice notes get the transcript in the caption, the rest as a reply. It uses the STT set up in Hermes. Open source, MIT. https://telesec.sergeybaus.com

    Tampilkan postingan
  • Magnus Hedemark@magnus919.com

    On Hermes Agent, I published a skill called Hermes-SkillOpt that will actually use a skill many times and find optimizations to improve its usefulness. Can't get that on an MCP.

    Tampilkan postingan
  • tef@tef.bsky.social

    my experience with machine translation tells me that people tend to build superstitions rather than intuition, and instead of better tools, we'll see "here's my special incantation for the context window" and "here's an elaborate pipeline with 20x token cost" with no real way of ever evaluating it

    Tampilkan postingan
  • Tom Aarsen@tomaarsen.com

    🤗 I've released Sentence Transformers v6.3 and wrote a blogpost to highlight its key new feature: token pooling-aware training! Train multi-vector models to preserve retrieval quality with fewer document vectors. Here's how 🧵 huggingface.co/blog/token-p...

    Tampilkan postingan
  • thebes@vgel.me

    new post: a WIP update on the chain lens, a new technique i've been working on for verbalizing model hidden states! the chain lens lets you extend single-token lenses like the J-lens into multi-token ones without any training, using an existing MTP head.

    Tampilkan postingan

Masuk untuk melihat lebih banyak

Akun yang mereka blokir

Blokir di Bluesky adalah catatan publik.

Akun ini tidak memblokir siapa pun.