Layer 8 · Neuron 6137 · boosted

L8N6137+1

Word quirks:

“ *”“ *“”“ *"”“ **"”“ "**”“ **“”“ "*”“ specific”

💤 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 artificial intelligence, a large language model trained by Google. My mission is to help people around the world by providing information, support, and solutions.

Specifically, my purpose is defined by two core principles:

To be helpful: I aim to assist users with tasks that are beneficial, whether they involve solving complex problems (like scientific research), translating languages, or generating creative content.
To do no harm: In cases where a task could have serious negative consequences (such as causing widespread misinformation or harm), I strive to ensure that the process is transparent, ethical, and safe

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. L13N6189-1💤	0.98	“ “”“ "”“ *”
2. L15N2709+1🕒	0.98	“ "”“ "”“ **“”
3. L12N214-1💤	0.98	“ "”“ "”“ **“”
4. L12N2725+1🕒	0.97	“ “”“ "”“ **“”
5. L14N6357-1💤	0.97	“ "”“ “”“ "*”
6. L14N835-1💤	0.97	“ “”“ "”“ *"”
7. L7N4561+1💤	0.97	“ "”“ "”“ ”
8. L6N1642-1🕒	0.97	“ ”“ "”“ *“”
9. L12N403-1🕒	0.96	“ ”“ "”“ *“”
10. L17N3442+1💤	0.96	“ "”“ "”“ "”

Shadow specimen, apart from the ten: L8N6137-1🕒, -0.14.

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 8, neuron 6137, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.82, top-1 agreement 0.63) 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
L8N6137+1 t0-v0.1	0.821	0.63	1.020	1.49	Top boosted tokens ↑
L8N6137-1🕒 shadow specimen t0-v0.1	1.474	0.59	1.752	1.50	“ punishment”“ penalty”“ penalties”“ punishments”“ penal”“ banning”“ restrictions”“ punish”

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.