specific method · filed under inference & serving
Effort-dependent exponential token-penalty schedule
The token-penalty coefficient decays exponentially with the effort setting, making lower effort levels penalize reasoning tokens more strongly.
Also called effort-dependent token-penalty coefficient, exponential token penalty schedule.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- core
How sources treat it
One count per evidence span, weakest treatment to strongest.
core 2
Documented in
Evidence
2 spans quoted from the sources, strongest treatment first.
During reinforcement learning, lower effort levels apply a stronger token penalty, whereas higher effort levels permit more computation.
coretraining objectivein DeepSeek-V4.1-FlashDeepSeek
The effort-dependent token-penalty coefficient is k(b)=k0 exp(−(b−bmin)/τ)
coretraining objectivein DeepSeek-V4.1-FlashDeepSeek
Filed alongside
Other methods under inference & serving :: reasoning control.
Configurable reasoning effortChain-of-thought reasoningDefault thinking modeDeployment-time scalar effort controlDisabling reasoning via chat-template configurationInference-time reasoning budget controlInterleaved thinking between tool callsThinking mode selectionCross-turn persistent reasoning historyMaximum thinking effortReasoning parserAlways-on thinking modeclear_thinking chat-template parameterControl-token-enabled thinking modeGenerate-verify-refine loopMedium-effort reasoning modeParallel-fewest-step samplingQwen3 soft thinking switchTask- and mode-specific sampling parameter recommendationsTest-time compute scalingAdaptive reasoningCapped linear reasoning-token length deductionConfigurable thinking or reasoning modeControllable thinking effort via system message and per-token cost