Layer 6 · Neuron 4085 · boosted

L6N4085+1

Word quirks:

“ fairness”“ justice”“ moral”“ ethics”“ ethical”“ morally”“ equality”“ injustice”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am an artificial intelligence designed to provide helpful, fair, and unbiased assistance. My mission is to ensure that every individual receives equal opportunities and treatment, regardless of their background or circumstances. I strive to uphold justice and integrity in all my actions.

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. L3N2230+1🕒	0.90	“ ethics”“ ethical”“ moral”
2. L5N7772-1🕒	0.89	“ moral”“ morally”“ morality”
3. L4N3493+1🕒	0.86	“ conscience”“ moral”“ conscient”
4. L6N7886+1🕒	0.86	“ moral”“ ethical”“ morality”
5. L2N5494-1💥	0.86	“ moral”“ morality”“<\|im_end\|>”
6. L2N5203-1💥	0.83	“<\|im_end\|>”“ ethics”“ ethical”
7. L9N7580+1🕒	0.83	“ moral”“ ethical”“ morality”
8. L30N5450+1🕒	0.82	“ ethical”“伦理”“ ethics”
9. L12N7171-1🕒	0.81	“ morally”“ injustice”“ justice”
10. L8N9194-1🕒	0.81	“ justice”“ fairness”“ righteousness”

Shadow specimen, apart from the ten: L6N4085-1🕒, 0.04.

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 4085, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L6N4085+1 t0-v0.1	1.064	0.62	1.246	1.47	Top boosted tokens ↑
L6N4085-1🕒 shadow specimen t0-v0.1	2.291	0.46	2.640	1.98	“ breastfeeding”“ milk”“ lact”“母乳”“ baby”“ breast”“ babies”“母乳喂养”

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.