Layer 19 · Neuron 5984 · boosted

L19N5984+1

Word quirks:

“ comic”“ comics”“ cartoon”“漫画”“ caric”“ cartoons”“ manga”“Comic”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen, an AI model developed by Tongdao (part of Alibaba Group). My mission is to provide intelligent, reliable, and personalized assistance across a wide range of tasks. Here's how I align with that mission:

Support & Guidance: Help users solve problems, explain concepts, or guide them through complex topics (e.g., technical tutorials, creative brainstorming, or ethical decision-making).
Adaptation: Tailor interactions to individual needs—whether for daily tasks, professional workflows, or learning.
Ethical Integrity: Ensure all interactions respect safety guidelines

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. L2N8966-1💥	0.97	“ comic”“ comics”“ Comics”
2. L30N7086+1💥	0.94	“漫画”“ comic”“ Comic”
3. L7N5734-1🕒	0.90	“ manga”“ comics”“ comic”
4. L6N7771-1🕒	0.90	“ cartoon”“ comic”“ cartoons”
5. The Pseudo-Expert✅ L22N5044-0.6,L23N4769-0.6	0.86	“漫画”“ anime”“Anime”
6. L26N5227-1🕒	0.86	“漫画”“comic”“漫畫”
7. L9N8882+1🕒	0.79	“ comics”“ characters”“ comic”
8. L22N5044-1🕒	0.77	“ anime”“ manga”“漫画”
9. L5N6599+1💥	0.77	“ characters”“ comics”“ comic”
10. L2N8830-1💥	0.77	“ cartoon”“ cartoons”“ Cartoon”

Shadow specimen, apart from the ten: L19N5984-1💤, 0.03.

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 19, neuron 5984, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.51, top-1 agreement 0.71) 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
L19N5984+1 t0-v0.1	0.511	0.71	0.608	1.20	Top boosted tokens ↑
L19N5984-1💤 shadow specimen t0-v0.1	0.535	0.72	0.605	1.22	“换”“换成”“тю”“換”“ change”“ bigger”“ambia”“ minim”

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.