Interface: ViterbiBiases

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:26

Six Viterbi transition biases matching the openai/privacy-filter viterbi_calibration.json schema (hosted at https://huggingface.co/software-mansion/react-native-executorch-privacy-filter-openai/blob/main/viterbi_calibration.json). Each value is added to the decoder score whenever the corresponding BIOES transition is taken. Positive values encourage the transition; negative values discourage it. Every field defaults to 0 (a neutral, validity-only Viterbi).

Properties

backgroundStay?

readonly optional backgroundStay: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:28

O -> O (background persistence). Higher = stay in background more, fewer false positives.

---

backgroundToStart?

readonly optional backgroundToStart: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:30

O -> B-x/S-x (span entry). Lower (negative) = enter spans more eagerly, higher recall.

---

endToBackground?

readonly optional endToBackground: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:32

E-x/S-x -> O (span closure to background).

---

endToStart?

readonly optional endToStart: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:34

E-x/S-x -> B-y/S-y (back-to-back spans).

---

insideToContinue?

readonly optional insideToContinue: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:36

B-x/I-x -> I-x (span continuation). Higher = longer spans.

---

insideToEnd?

readonly optional insideToEnd: number

Defined in: extensions/nlp/utils/privacyFilterUtils.ts:38

B-x/I-x -> E-x (span closure). Higher = shorter spans.