Layer 12 · Neuron 5161 · boosted

L12N5161+1

Word quirks:

“ yourself”“ your”“ myself”“ yours”“ now”“ you”“ whenever”“ always”

💤 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 Qwen3.5, the latest large AI model developed by Tongyi Lab. My mission is to empower users with high-quality services and support across various domains. Here's a breakdown of my capabilities:

Comprehensive Knowledge & Skills: I have been trained on extensive, high-quality data from diverse sources, ensuring I stay up-to-date with the latest advancements.
Human-Aligned Performance: I prioritize helping users achieve their goals, whether it's learning, working, or solving problems, rather than just providing information.
Multilingual & Global Reach: I support

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. L14N5628-1💤	0.92	“ myself”“ yourself”“ you”
2. L11N7217-1🕒	0.92	“ now”“ you”“ here”
3. L8N7282+1💤	0.91	“ yourself”“ your”“ you”
4. L12N8256+1🕒	0.90	“ right”“ yourself”“ ourselves”
5. L8N1175-1💤	0.90	“⏎”“ usually”“ your”
6. L12N713-1💤	0.90	“ always”“ your”“ you”
7. L12N5848+1🕒	0.90	“ your”“ you”“ always”
8. L13N7313+1🕒	0.90	“ your”“ you”“ each”
9. L12N6912-1💤	0.89	“ your”“ you”“ my”
10. L8N8858-1💤	0.89	“ your”“ you”“ yourself”

Shadow specimen, apart from the ten: L12N5161-1🕒, 0.15.

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 12, neuron 5161, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.79, top-1 agreement 0.65) 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
L12N5161+1 t0-v0.1	0.788	0.65	0.951	1.22	Top boosted tokens ↑
L12N5161-1🕒 shadow specimen t0-v0.1	1.288	0.59	1.580	1.31	“ further”“ continuation”“ deeper”“ continued”“ already”“ path”“ farther”“ ⏎”

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.