Layer 3 · Neuron 8986 · suppressed

L3N8986-1

Word quirks:

“⏎”“ ‘”“ generated”“ generation”“ reference”“ referenced”“ context”“ generating”

💤 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, a large language model. I do not have a physical body or consciousness. I exist to process information, generate text, and assist with tasks. I am here to answer your question.

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. L3N2492+1💤	0.91	“⏎”“ generation”“ “”
2. L5N3516+1💤	0.89	“ generated”“ generating”“ generation”
3. L7N2566-1💤	0.88	“ “”“ referenced”“ ‘”
4. L6N6256+1💤	0.87	“ ‘”“ generation”“ referenced”
5. L5N1601+1💤	0.87	“ generated”“ reference”“ ‘”
6. L4N5713-1💤	0.85	“⏎”“ ‘”“ referenced”
7. L5N6769-1🕒	0.85	“ generated”“ ‘”“ “”
8. L5N8018-1💤	0.85	“ generated”“ generating”“ generation”
9. L3N7703+1💤	0.84	“⏎”“ referenced”“ ‘”
10. L5N885+1💤	0.83	“ referenced”“ generating”“⏎”

Shadow specimen, apart from the ten: L3N8986+1🕒, -0.11.

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 3, neuron 8986, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.88, top-1 agreement 0.62) 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
L3N8986+1🕒 shadow specimen t0-v0.1	2.278	0.42	2.699	2.52	“ chicken”“ chickens”“ poultry”“ Chicken”“ meat”“<\|im_end\|>”“ cooked”“ seafood”
L3N8986-1 t0-v0.1	0.878	0.62	1.055	1.62	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.