Liquid AI launches 230M and 350M encoders that stay fast on long context
helloiamleonie · x · 2026-07-29
Liquid AI says its new LFM2.5-Encoders are tiny, fast bidirectional encoders built from the LFM2.5 backbones. The company replaced causal attention with bidirectional attention, made the short convolutions non-causal, and trained with a masked-language-modeling objective. Two models are released: LFM2.5-Encoder-230M and LFM2.5-Encoder-350M.
The release targets classification, retrieval, intent routing, safety filters, and other encoder-based workloads that often run on CPUs. Liquid AI says the models stay efficient at long context lengths up to 8,192 tokens and can encode a 12–15 page document, about 8k tokens, in under 30 seconds on a CPU.
Related event: Liquid AI releases LFM2.5 Encoder 230M and 350M(7 posts)→
More from Embodied
- Sentdex trains a fall-recovery policy for the Asimov humanoid: getting up is hard — Sentdex · 2026-10-16
- LeRobot adds VLM-powered feature that turns plain-language goals into robot skills — Gradio · 2026-10-16
- Optimus Robot Shows Off Its Driving in a Tesla Roadster — debreuil · 2026-10-16
- Qdrant Edge gives the Reachy Mini robot local persistent semantic memory — tomaarsen · 2026-10-16
- Tesla Roadster Thruster Teardown: SpaceX COPV Tank Plus Optimus Actuator — XFreeze · 2026-10-16
- Critic Calls Figure's Terminator-Style Robot Demo Publicity Over Substance — FuSheng_0306 · 2026-10-16