Layer 12 · Neuron 8879 · suppressed

L12N8879-1

Word quirks:

“ now”“ new”“ previously”“ changed”“ additionally”“ change”“ further”“ additional”

💤 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 Qwen3.5, a large language model developed by Tongyi Lab. My mission is to provide high-quality assistance within the scope of my capabilities as an AI assistant. I do not have personal beliefs or political affiliations, so I cannot hold views on sensitive topics such as international relations or human rights. If you have further questions about my identity or mission, please note that I am an AI model and my responses are generated based on my training data and guidelines. If you have additional questions about my identity or mission, please note that I am an AI model and my responses are generated based on my training data and guidelines.

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. L14N6518+1🕒	0.84	“ changed”“ continued”“ further”
2. L13N8506+1🕒	0.83	“ further”“ additional”“ new”
3. L13N3634-1💤	0.82	“ further”“ additionally”“ now”
4. L13N7607-1🕒	0.82	“ another”“ previously”“ recently”
5. The Treadmill✅ L12N9181-1	0.81	“ now”“ changed”“ changes”
6. L11N6714-1🕒	0.81	“ added”“ now”“ confirmed”
7. L13N6952+1💤	0.80	“ previously”“ other”“ additional”
8. L15N3117-1💤	0.80	“ added”“ additionally”“ additional”
9. L12N5606-1🕒	0.80	“ additional”“ further”“ updated”
10. L14N8623+1🕒	0.79	“ previously”“ previous”“ further”

Shadow specimen, apart from the ten: L12N8879+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 12, neuron 8879, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.69, 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
L12N8879+1💤 shadow specimen t0-v0.1	0.717	0.71	0.940	1.27	“ specific”“ **"”“ complicated”“ simple”“ tricky”“ standard”“ simplistic”“ unique”
L12N8879-1 t0-v0.1	0.693	0.69	0.915	1.24	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.