Layer 4 · Neuron 6815 · boosted

L4N6815+1

Word quirks:

“ ‘”“ '”“<|im_end|>”“⏎”“ \"”“!”“ !”“ `”

💤 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 artificial intelligence assistant designed to provide helpful, harmless, and non-deceptive responses. My mission is to assist users with answering their questions, providing information, and offering assistance in a variety of tasks. I strive to be helpful, harmless, and non-deceptive in my interactions with users.

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. L3N6190+1🕒	1.00	“⏎”“!”“ \"”
2. L4N15+1💤	0.99	“ ‘”“<\|im_end\|>”“ '”
3. L3N1814+1🕒	0.98	“ ‘”“⏎”“<\|im_end\|>”
4. L4N346+1💤	0.98	“ ‘”“!”“<\|im_end\|>”
5. L4N3207-1🕒	0.98	“⏎”“<\|im_end\|>”“ \"”
6. L3N7053+1🕒	0.97	“ '”“ ‘”“<\|im_end\|>”
7. L3N1202+1🕒	0.97	“<\|im_end\|>”“ \"”“ ‘”
8. L3N4549+1🕒	0.97	“ ‘”“!”“ '”
9. L4N749-1🕒	0.97	“<\|im_end\|>”“ ‘”“ !”
10. L3N1279+1🕒	0.96	“⏎”“ ⏎”“ !”

Shadow specimen, apart from the ten: L4N6815-1🕒, -0.18.

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 4, neuron 6815, 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 0.66) 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
L4N6815+1 t0-v0.1	0.776	0.66	0.962	1.46	Top boosted tokens ↑
L4N6815-1🕒 shadow specimen t0-v0.1	1.760	0.51	2.085	2.04	“ French”“ France”“ french”“ organisation”“ organisations”“ centres”“ Paris”“<\|im_end\|>”

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.