Unverified50% confidenceOpinionExact time
Boris described dynamic workflow as the 'fourth scaling factor' after model size, data volume, and training compute.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Claude Code Lead: Since Opus 4.5, 100% of My Code Is Written by AI
bilibili尽我所能HH10/16/2026
Related Entities
Related Claims
UnverifiedClaude 4's extended reasoning can be combined with tool use, and the model is able to reason about when to call functions, interpret results, reason again, and coherently chain actions over long spans.66% similarUnverifiedThis data-free training scheme achieves performance comparable to a diffusion model trained with HMC reference data in two-dimensional φ⁴ theory.65% similarUnverifiedTraining large models with backpropagation requires storing the global computation graph (activation cache), and GPU memory consumption is typically 3–4 times or more that of the inference phase.65% similarUnverifiedThe paper argues that assertions about hosted model behavior must be explicitly annotated along four dimensions: the model identifier tested, the service period, the measurement tool, and the inference configuration.65% similarUnverifiedMerging the parameters of multiple trained models to generate new trade-off combinations has been shown to be helpful.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1058743API
curl https://kongchang.com/api/v1/knowledge/claims/1058743MCP
get_claim(id=1058743)