Description
Compose a positive logit scale with two temperatures.
Log probabilities, logits, guidance scales and sequence scores use different conventions. Temperature changes logit scale, an additive bias changes odds, and a sequence-length penalty changes ranking. One supplied scalar contribution can be checked independently of a complete probability distribution or model-generation process.
Compose a positive logit scale with two temperatures.
When to use this tool
Use Logit Scale Temperature Composition when checking the stated probability, score, guidance or scale contribution.
How to use the tool and how it works
- Enter scale, temperature, extraTemperature. Use the units and conventions below.
- Run the tool. Correct any format, range, or cross-field input error before interpreting the output.
- Read effectiveScale, equivalentTemperature. Use the output convention below.
Compose a positive logit scale with two temperatures. Require all three factors > 0; effectiveScale = scale/temperature/extraTemperature, equivalentTemperature = temperature*extraTemperature/scale.1
| Input field | Meaning and units | Accepted input |
|---|---|---|
| scale | Scale | Required. Finite number; strictly above 0. |
| temperature | Temperature | Required. Finite number; strictly above 0. |
| extraTemperature | Extra Temperature | Required. Finite number; strictly above 0. |
| Output field | Interpretation | Returned shape |
|---|---|---|
| effectiveScale | Effective Scale | Required. Finite number. |
| equivalentTemperature | temperature*extraTemperature/scale on the helper’s strictly positive scale convention. | Required. Finite number. |
{
"input": {
"scale": 2,
"temperature": 4,
"extraTemperature": 0.5
},
"output": {
"effectiveScale": 1,
"equivalentTemperature": 1
}
}Limitations and assumptions
- This scalar equivalence does not apply softmax, fit calibration or guarantee unchanged sampling after other logit operations.
- Inputs must satisfy both the field types and the cross-field conditions above. Invalid input causes an error; a successful false flag or documented placeholder is a different result.
- Numeric calculations use binary floating point unless exact decimal-integer arithmetic is explicitly stated. No hidden measurement tolerance or decimal rounding is added; nonfinite or out-of-contract results are rejected. Numeric signed-zero results are exported as zero.
Alternative or Complementary methods
Use the full distribution and model-specific decoding or training objective for normalized probabilities, preference optimization or actual generation.
References
Similar or alternative tools
- Logit Bias One Token Odds Factor
Return one-token logit-bias odds transformation.
- Log Probability Nats Bits Crosswalk
Convert a nonpositive log probability from nats to bits.