Layer 14 · Neuron 2123 · suppressed

L14N2123-1

Word quirks:

“ analogous”“ similar”“ thus”“ similarly”“⏎”“ here”“ basically”“ reflected”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

As an AI model developed by Google, I am a large language model designed to assist with a wide range of tasks. My primary function is to provide helpful, accurate, and safe information. Whether you need help solving a problem, generating ideas, or just want to chat, I'm here to support you. If you have questions or need assistance, feel free to ask! 😊

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. L13N4629-1🕒	0.90	“ similarly”“ analogous”“ thus”
2. L15N531-1🕒	0.88	“ similar”“ comparable”“ similarly”
3. L12N1905+1🕒	0.88	“ similar”“ analogous”“ similarly”
4. L13N4452+1💤	0.87	“ similar”“ similarly”“ thus”
5. L16N5205+1💤	0.86	“ similar”“ thus”“ below”
6. L13N2378+1💤	0.86	“ similar”“ similarly”“ analogous”
7. L14N6113-1🕒	0.86	“ similar”“⏎”“ similarly”
8. L14N8186-1🕒	0.86	“ similar”“ similarly”“ basically”
9. L13N4451+1🕒	0.86	“ basically”“ similar”“ __”
10. L17N6118-1🕒	0.85	“ similar”“ similarly”“ yourself”

Shadow specimen, apart from the ten: L14N2123+1💤, 0.10.

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 14, neuron 2123, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.78, top-1 agreement 0.65) 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
L14N2123+1💤 shadow specimen t0-v0.1	0.700	0.65	0.953	1.31	“ what”“ ⏎”“ my”“ ___”“ __”“ consistently”“�”“ always”
L14N2123-1 t0-v0.1	0.778	0.65	0.954	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.