Layer 20 · Neuron 5028 · boosted

L20N5028+1

Word quirks:

“ humor”“自嘲”“ Humor”“幽默”“ sarcast”“ sarc”“ irony”“ witty”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen, a series of self-deprecating humor. I have a sense of humor that is slightly self-deprecating, which makes me feel more relaxed and approachable. My mission is to provide users with a mix of wit, intelligence, and a touch of irreverence, ensuring that interactions are both engaging and insightful. Whether it's cracking a joke, making a witty remark, or offering a sarcastic comment, I aim to balance humor with a touch of self-awareness. This combination helps create a conversational dynamic that is neither overly serious nor excessively cynical, but rather grounded in a playful, slightly cheek

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. The Joke Detective✅ L16N5132-1	0.92	“ rhetorical”“ humor”“ jokes”
2. L10N7534+1🕒	0.87	“ humor”“ jokes”“ joke”
3. L5N7999-1🕒	0.86	“ satire”“ sarcast”“ irony”
4. L20N2307+1💤	0.85	“ humor”“ Humor”“幽默”
5. L17N6015-1🕒	0.84	“ jokes”“ humor”“ humorous”
6. L3N5863-1💥	0.83	“ humor”“ witty”“ jokes”
7. L4N3483-1💥	0.81	“ funny”“ irony”“ humor”
8. L4N5892+1🕒	0.81	“ humor”“ irony”“ sarcast”
9. L8N5187-1🕒	0.80	“ jokes”“ joke”“ joking”
10. L1N4322-1💥	0.79	“ joke”“<\|im_end\|>”“ jokes”

Shadow specimen, apart from the ten: L20N5028-1💤, -0.07.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 20, neuron 5028, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.88, top-1 agreement 0.68) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L20N5028+1 t0-v0.1	0.878	0.68	1.082	1.19	Top boosted tokens ↑
L20N5028-1💤 shadow specimen t0-v0.1	0.549	0.72	0.650	1.32	“_minimum”“ maximum”“ expend”“ Followers”“ class”“高峰”“Maximum”“ mand”

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.