Layer 6 · Neuron 7620 · boosted

L6N7620+1

Word quirks:

“ ____”“ ______”“ _____”“ __”“ ________”“ ___”“ _”“ __________________”

💤 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 AI assistant designed to help with technical tasks. My mission is to support users in solving problems related to computer science, mathematics, and engineering. If you have a specific task or project you'd like assistance with, feel free to ask!

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. L15N890-1💤	0.99	“ ____”“ __”“ ___”
2. L17N1276-1💤	0.99	“ ___”“ __”“ ______”
3. L10N1857-1🕒	0.99	“ _____”“ ______”“ ____”
4. L10N4326-1🕒	0.99	“ __”“ ________”“ ____”
5. L9N3505-1🕒	0.99	“ ______”“ ____”“ ________”
6. L15N4110-1🕒	0.99	“ __”“ ____”“ _____”
7. L11N2536+1🕒	0.99	“ __”“ ____”“ ___”
8. L14N3433+1🕒	0.99	“ ____”“ __”“ ______”
9. L11N623+1🕒	0.99	“ ____”“ ______”“ ________”
10. L11N68+1🕒	0.99	“ __”“ ____”“ _____”

Shadow specimen, apart from the ten: L6N7620-1💥, -0.20.

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 6, neuron 7620, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.98, top-1 agreement 0.66) 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
L6N7620+1 t0-v0.1	0.976	0.66	1.200	1.56	Top boosted tokens ↑
L6N7620-1💥 shadow specimen t0-v0.1	2.734	0.42	3.089	2.11	“ cards”“塔罗”“ tar”“ Tar”“ card”“ Cards”“ divine”“-card”

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.