Search
Have an existing account?
Sign In
Tag: long context
A one-researcher model lab ships a 10M-token coding engine
Mirendil says Fern 0.1 was trained for under $10M with agents writing more than 99 percent of the…
Stanford and Berkeley shrink the first pass over long prompts
TokenGate learns which tokens matter and drops the rest before attention runs in full.
DeepSeek shrinks its KV cache to fit million-token contexts
DeepSeek V4.1 Flash targets the memory bill behind long-running agents, cutting cache footprint to 890 bytes per token.
Pokee’s single-GPU model opens a ten-million-token window
Regulated industries get cloud-class long context without sending data anywhere.