Encoders strike back: Jev bets agents need classifier-style decision efficiency
dotey · x · 2026-09-19
- Encoder路线 (BERT) was efficient at compressing inputs but stuck in task-specific fine-tuning; decoder-only won via scaling, ICL and instruction tuning, turning everything into generation.
- Agents change the workload: huge inputs, heavy KV caches, frequent looped calls — full prefill plus autoregressive decoding of a few tokens per judgment is getting expensive.
- Jev aims to complete BERT's unfinished path: keep GPT-style task generalization while regaining encoder/classifier decision efficiency, as a generalized decision model for agents and services.
More from Models
- Codex usage limits finally reset after days of users hitting exhausted quotas — CtrlAltDwayne · 2026-09-19
- Rumored Opus 5.2 generates a full 5-minute Titanic film in pure JavaScript with three.js — dotey · 2026-09-19
- Model Remembers 'cranberry-42' Across App Restarts in Persistent-Session Test — RileyRalmuto · 2026-09-19
- Free models aren't the real problem: studies show hallucinations persist in SOTA LLMs — AryHHAry · 2026-09-19
- Why do AI models lack personality? Safety training and attachment fears, explained — alexisgallagher · 2026-09-19
- New model JEV is free on OpenRouter, Vercel, Cloudflare and others, at ~$0.042/M input tokens — airesearch12 · 2026-09-19