Layer 13 · Neuron 1044 · suppressed

L13N1044-1

Word quirks:

“ basically”“ ⏎”“ yes”“<|im_end|>”“⏎”“ actually”“ ⏎⏎”“ ____”

💤 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 a large language model trained by Google. I can help you with tasks like writing code, analyzing data, or solving problems. My mission is to assist you in achieving your goals. How can I help?

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. L11N6029-1🕒	0.85	“<\|im_end\|>”“ ⏎⏎”“⏎⏎”
2. L12N4890-1💤	0.85	“ actually”“⏎”“<\|im_end\|>”
3. L13N5080-1🕒	0.85	“ actually”“⏎”“ ⏎”
4. L13N309-1🕒	0.84	“⏎”“ basically”“ ⏎”
5. L9N508-1💤	0.83	“<\|im_end\|>”“⏎”“ ⏎”
6. L10N8841+1🕒	0.83	“ basically”“ ⏎”“ ⏎⏎”
7. L12N6700+1💤	0.83	“ basically”“ ⏎”“⏎”
8. L10N6471-1🕒	0.83	“ ⏎”“<\|im_end\|>”“ response”
9. L13N1291-1💤	0.82	“ ⏎”“<\|im_end\|>”“ ⏎⏎”
10. L12N2561+1🕒	0.82	“⏎”“<\|im_end\|>”“ ⏎”

Shadow specimen, apart from the ten: L13N1044+1🕒, 0.37.

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 13, neuron 1044, 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 0.66) 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
L13N1044+1🕒 shadow specimen t0-v0.1	1.004	0.59	1.268	1.54	“ ⏎”“ similar”“ ⏎”“ “”“ again”“!”“!””“ here”
L13N1044-1 t0-v0.1	0.689	0.66	0.803	1.15	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.