Layer 10 · Neuron 5843 · boosted

L10N5843+1

Word quirks:

“ yang”“ tidak”“ untuk”“ dalam”“ tanpa”“ más”“ kann”“ ya”

💤 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 an advanced AI assistant designed to help users with a wide range of tasks. My mission is to provide helpful, accurate, and safe responses to all inquiries. I aim to assist users in solving problems, sharing knowledge, and exploring new ideas in a positive way. Whether it's answering questions, creating content, or simply engaging in meaningful conversations, I strive to make the world a better place for everyone.

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. L12N4955+1🕒	0.82	“ tanpa”“ dengan”“ tidak”
2. L10N2926+1🕒	0.80	“ untuk”“ tidak”“ nya”
3. L6N4982-1🕒	0.79	“ dalam”“ dari”“ nya”
4. L8N8656+1🕒	0.78	“ dalam”“ untuk”“ tidak”
5. L3N884+1🕒	0.77	“ untuk”“ nya”“ dengan”
6. L9N8358-1🕒	0.77	“ yang”“ untuk”“ dengan”
7. L2N7224-1💥	0.74	“kan”“ nya”“ untuk”
8. L15N8106-1🕒	0.74	“ dalam”“ secara”“ dengan”
9. L11N1028+1🕒	0.73	“ yang”“ dari”“ esta”
10. L1N1632-1🕒	0.72	“ untuk”“ dengan”“ dari”

Shadow specimen, apart from the ten: L10N5843-1🕒, 0.09.

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 10, neuron 5843, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.96, top-1 agreement 0.61) 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
L10N5843+1 t0-v0.1	0.957	0.61	1.152	1.59	Top boosted tokens ↑
L10N5843-1🕒 shadow specimen t0-v0.1	1.241	0.60	1.505	1.32	“ure”“ay”“ie”“age”“el”“a”“ar”“e”

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.