Layer 3 · Neuron 5080 · boosted

L3N5080+1

Word quirks:

“ parent”“ parents”“ Parent”“ families”“ Parents”“ family”“-parent”“ parenting”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

Hello! 👋
Thank you for reaching out to our school community. We're here to support your child's journey with care and dedication. 🌟📚

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. L10N7414-1🕒	0.98	“ parents”“ parental”“ parent”
2. L5N8495-1🕒	0.98	“ parents”“ parent”“ parenting”
3. L9N7405+1🕒	0.97	“ parent”“ parents”“ parental”
4. L19N7874+1🕒	0.97	“ parental”“ parent”“ parents”
5. L10N2730-1🕒	0.97	“ parent”“ parents”“ parental”
6. L18N9125+1💤	0.96	“ parents”“ Parents”“Parents”
7. L8N3647+1🕒	0.96	“ parents”“ parent”“ parental”
8. L11N3900-1🕒	0.96	“ parents”“ parent”“ family”
9. L11N3100+1🕒	0.96	“ parental”“ parents”“ parent”
10. L2N4444+1💥	0.96	“ parent”“ Parent”“ parents”

Shadow specimen, apart from the ten: L3N5080-1💤, -0.12.

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 3, neuron 5080, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 3.01, top-1 agreement 0.38) 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
L3N5080+1 t0-v0.1	3.014	0.38	3.573	2.47	Top boosted tokens ↑
L3N5080-1💤 shadow specimen t0-v0.1	0.834	0.62	1.026	1.59	“⏎”“ words”“ ‘”“ \"”“ code”“ name”“⏎ ⏎”“ ””

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.