← all emotions · markdown (agent skill)

Pride conditioning guide

Target: EmoNet Pride strength. reward = (genu + blend + 1.25·emo)/(1+WER).

TL;DR. Best reward comes WITHOUT the LoRA, and in fact without even the steering prompt — plain BASE (0.489) beats the evolved prompt (BASE_P 0.437) and every merge, which decline monotonically (LoRA50 0.403 to LoRA100 0.319). Pride is a failure case in this rig: both prompt-steering and the LoRA reduce reward, and emo never rises (max 0.018 at LoRA50), so keep a plain neutral prompt and leave both LoRA and steering prompt off. The LoRA's only real effect is to erode vocal-burst blend (0.453 to 0.263) for zero emotional payoff. When forced, its coloring is counterproductive: outputs correlate with Triumph, Contempt and Malevolence, collapsing dignified pride into an arrogant, contemptuous tone.

Without a LoRA

Best: neutral prompt, no LoRA → reward 0.49
GENERAL: A voice overwhelmingly expressing pride, a proud, self-assured voice, chest swelling with pride, triumphantly boastful, impossible to hide. SCRIPT: (with swelling, self-assured pride) "<neutral sentence>" · temp 1.05 top_p 0.95 top_k 25

With the emotion LoRA (dose)

50% ★ — reward 0.40 · emo 0.02
genu 0.11 · blend 0.34 · qual 0.58 · WER 0.30
100% — reward 0.32 · emo 0.00
genu 0.11 · blend 0.28 · qual 0.53 · WER 0.36
150% — reward 0.33 · emo 0.01
genu 0.12 · blend 0.26 · qual 0.52 · WER 0.28

Conditions

conditionrewardemotiongenublendqualityWER
neutral prompt, no LoRA0.490.000.110.450.630.27
evolved prompt, no LoRA0.440.000.120.410.580.38
evolved prompt + LoRA 50%0.400.020.110.340.580.30
evolved prompt + LoRA 100%0.320.000.110.280.530.36
evolved prompt + LoRA 150%0.330.010.120.260.520.28

Side effects & correlations (LoRA @ 100%)

↑ pushed up: Impatience and Irritability (+0.29), Fatigue Exhaustion (+0.19), Affection (+0.12)
↓ pushed down: Concentration (-0.36), Fullness (-0.21), Vocal-burst blend (-0.17)
+ correlates: Malevolence Malice (+0.68), Contempt (+0.59), Amusement (+0.49)
− anti-correlates: Chunking (-0.33), Focus (-0.29), Esthetics (-0.29)

Conclusion

Best reward comes WITHOUT the LoRA, and in fact without even the steering prompt — plain BASE (0.489) beats the evolved prompt (BASE_P 0.437) and every merge, which decline monotonically (LoRA50 0.403 to LoRA100 0.319). Pride is a failure case in this rig: both prompt-steering and the LoRA reduce reward, and emo never rises (max 0.018 at LoRA50), so keep a plain neutral prompt and leave both LoRA and steering prompt off. The LoRA's only real effect is to erode vocal-burst blend (0.453 to 0.263) for zero emotional payoff. When forced, its coloring is counterproductive: outputs correlate with Triumph, Contempt and Malevolence, collapsing dignified pride into an arrogant, contemptuous tone.