Unverified50% confidenceFactExact time
The research team used 20 LLMs to generate 44,782 stories in four languages from textual and visual stimuli, involving a total of 88,919 named characters.
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe research team sampled 200,000 grounding examples from DeskForge-1M to fine-tune four vision-language models.65% similarUnverifiedA text token roughly corresponds to 0.75 English words or approximately 1.5 Chinese characters, and image visual encoders typically split a single image into hundreds of visual tokens.64% similarUnverifiedThe research team compressed the original 468,000-sentence source text into a streamlined annotation collection.60% similarUnverifiedThe developer defined "recovery-type" sessions as 10 in total with 1,636 messages combined, and "publishing-type" sessions as 3 in total with 536 messages combined; the 1,327-message session was thrown into "uncategorized" because it could not be assigned to either of these two categories.60% similarUnverified上下文窗口决定模型单次能处理的文本量：8K约6000词，32K约24000词，128K可处理整本书内容60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1129024API
curl https://kongchang.com/api/v1/knowledge/claims/1129024MCP
get_claim(id=1129024)