The gap
Run this through the current pipeline:
"Hey! Terraform apply should be run from CI, not by you."
tokens : hey terraform apply should be run from ci not by you
lexicon hits : `not` → frustration/negation ← the only hit
paralinguistic : (none — one `!` is not repeated punctuation, `Hey!` is not shouting)
turn-pair-core : correction, implicit (regex `should\s+be`)
The lexicon contributes exactly one token, and it is the single most generic in the corpus (not, 2,794 hits). Remove that word and the lexicon says nothing at all.
Yet this is close to ideal proposal material: a corrective directive with a generalisable rule — terraform apply belongs to CI, not to you — which converts almost directly into a standing instruction. And the speaker is completely calm.
Why the lexicon misses it, correctly
The prompt asks: "does this entry express how the person FEELS about the interaction?" By that test every token here is genuinely neutral, and terraform/apply/ci are tool names the prompt was explicitly taught (#56) to neutralise. The lexicon is working as designed. The design measures the wrong dimension for this class.
frustration | praise | neutral is an affect axis. This utterance has near-zero affect and high corrective force. Those are independent.
This may mean #56 pruned useful signal
In #56 I classified must (706 hits) and do (1,620) as noise and taught the prompt to neutralise reporting verbs. But must and should are precisely the markers of this class — "you must not", "should be run from CI", "you should have edited the body".
turn-pair-core's own regex already agrees: its WEAK patterns include should\s+(use|do|be|have) and please\s+(use|do|try|don't), on the theory that directives signal correction. The deterministic layer models this; the learned lexicon does not.
This should be measured before acting. Sample the existing must/should/do hits and classify how many are genuine directives versus ordinary prose. I do not know the split, and the fix differs depending on it.
Proposed shape
Give a lexicon entry two independent fields rather than one:
- polarity (affect):
frustration | praise | neutral — unchanged
- speech act:
directive | corrective | report | none
A term can be neutral in feeling and strongly corrective in force: should, must, instead, always, never, don't. The second axis is probably the better predictor of a useful proposal, since a calm correction is exactly what converts into a standing instruction.
Downstream, turn-frustration would emit the speech-act dimension alongside polarity, and session-overview could weight corrective turns distinctly from angry ones.
Acceptance
Related
- Companion to the signal-modulator issue:
! amplifies a directive as readily as it amplifies anger.
The gap
Run this through the current pipeline:
The lexicon contributes exactly one token, and it is the single most generic in the corpus (
not, 2,794 hits). Remove that word and the lexicon says nothing at all.Yet this is close to ideal proposal material: a corrective directive with a generalisable rule — terraform apply belongs to CI, not to you — which converts almost directly into a standing instruction. And the speaker is completely calm.
Why the lexicon misses it, correctly
The prompt asks: "does this entry express how the person FEELS about the interaction?" By that test every token here is genuinely neutral, and
terraform/apply/ciare tool names the prompt was explicitly taught (#56) to neutralise. The lexicon is working as designed. The design measures the wrong dimension for this class.frustration | praise | neutralis an affect axis. This utterance has near-zero affect and high corrective force. Those are independent.This may mean #56 pruned useful signal
In #56 I classified
must(706 hits) anddo(1,620) as noise and taught the prompt to neutralise reporting verbs. Butmustandshouldare precisely the markers of this class — "you must not", "should be run from CI", "you should have edited the body".turn-pair-core's own regex already agrees: itsWEAKpatterns includeshould\s+(use|do|be|have)andplease\s+(use|do|try|don't), on the theory that directives signal correction. The deterministic layer models this; the learned lexicon does not.This should be measured before acting. Sample the existing
must/should/dohits and classify how many are genuine directives versus ordinary prose. I do not know the split, and the fix differs depending on it.Proposed shape
Give a lexicon entry two independent fields rather than one:
frustration | praise | neutral— unchangeddirective | corrective | report | noneA term can be neutral in feeling and strongly corrective in force:
should,must,instead,always,never,don't. The second axis is probably the better predictor of a useful proposal, since a calm correction is exactly what converts into a standing instruction.Downstream,
turn-frustrationwould emit the speech-act dimension alongside polarity, andsession-overviewcould weight corrective turns distinctly from angry ones.Acceptance
must/should/dohits.not.ci,gh,pr,re,non,but,pushed) stay neutral.frustration-lexicon, with old verdicts preserved byreviseslineage as in fix(lexicon): keep hyphenated compounds whole, and stop flagging ordinary grammar #56.Related
!amplifies a directive as readily as it amplifies anger.