Layer 9 · Neuron 8181 · suppressed

L9N8181-1

Word quirks:

“ now”“ transformed”“ changed”“ converted”“ actually”“ better”“ “”“ updated”

💤 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 developed by Google. My mission is to provide helpful, safe, and ethical responses to your questions while adhering to my safety guidelines. I don't have personal feelings or a physical form, but I'm here to assist you with information and problem-solving tasks.

If you'd like to explore more about my capabilities or discuss how I can help you, feel free to share your specific needs! 😊

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. L11N6333+1💤	0.86	“ now”“ changed”“ previously”
2. Great News✅ L12N7152-1	0.83	“ now”“ changed”“ change”
3. L15N7258-1🕒	0.81	“ changed”“ converted”“ replaced”
4. L12N4980+1🕒	0.81	“ changed”“ transformed”“ ⏎”
5. L9N4342+1💤	0.81	“ transition”“ previous”“ now”
6. L10N4702+1🕒	0.81	“ changed”“ now”“ ”
7. L9N5198-1🕒	0.79	“ changed”“ replaced”“ modified”
8. L8N5728+1💤	0.78	“ now”“ changed”“ old”
9. L16N3123-1🕒	0.78	“ changed”“ “”“ changing”
10. L13N8824-1🕒	0.78	“ previously”“ corrected”“ revamped”

Shadow specimen, apart from the ten: L9N8181+1🕒, 0.08.

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 9, neuron 8181, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.55, top-1 agreement 0.69) 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
L9N8181+1🕒 shadow specimen t0-v0.1	1.552	0.50	1.909	1.92	“ integration”“ seperate”“ integrated”“ integrate”“ separate”“ Integration”“ integrates”“ main”
L9N8181-1 t0-v0.1	0.554	0.69	0.723	1.29	Top boosted tokens ↑

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.