Unverified50% confidenceBenchmarkExact time
ChunkAct reduces input tokens by 4.4x and output tokens by 2.3x on LIBERO-Pro
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/14/2026
First Seen
Valid until: 1/12/2027
Sources
Related Entities
Related Claims
UnverifiedThe embedding model's maximum length means that inputs exceeding 512 tokens are silently truncated, so the example sets the chunk budget to 480 tokens.61% similarUnverifiedREMORY's residual decoder generates a bounded soft token sequence, each a 768-dimensional vector, stopping after N ≤ 64 tokens.61% similarUnverifiedThe same code refactoring task may consume 3,000 tokens in standard mode and may balloon to over 50,000 tokens in Extra High mode.60% similarUnverifiedWhen 32 requests share the same 4k-token system prompt, Superfluid's prefix cache hit rate reaches 97%, on par with Ollama, while mlx_lm.server is 0%.60% similarUnverifiedOn the same machine and model, the decoding speed difference between low power mode and high power mode is 1.4 to 3.8 times; for example, running Qwen 3.5 35B on Ollama gives 51.9 tokens/s in low power and 107.1 tokens/s in high power.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1028137API
curl https://kongchang.com/api/v1/knowledge/claims/1028137MCP
get_claim(id=1028137)