Unverified50% confidenceFactExact time
Response ratio is defined as the number of answer tokens as a proportion of the sample's total token count.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe input/output ratio refers to the ratio between the number of input tokens consumed and the number of output tokens generated in a single API call; in regular chat scenarios, this ratio typically ranges from single digits to a few dozen to one.64% similarUnverifiedIn this usage scenario, two-thirds of the token volume comes from requests with contexts exceeding 200,000 tokens, while requests below 50,000 tokens contribute only 2.5% of the token volume.62% similarUnverifiedIn the control probe, when the 128 token limit was sent nested under the extra_body parameter, the model returned empty responses; after passing max_tokens as a top-level parameter, all 200 responses were produced normally.61% similarUnverifiedRefusal rate is a core quantitative metric in AI safety evaluation, referring to the proportion of requests in a specific category that a model explicitly declines to execute or avoids answering, usually tallied by human annotation or an automatic judge.60% similarUnverifiedHit rate usually refers to the proportion of queries whose results contain at least one relevant document; top-1 hit rate requires the most relevant result to rank first, and top-5 hit rate allows relevant results to appear in the top five.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102337API
curl https://kongchang.com/api/v1/knowledge/claims/1102337MCP
get_claim(id=1102337)