Layer 5 · Neuron 3937 · boosted

L5N3937+1

Word quirks:

“!”“"!”“ !”“!:”“!\”“!"”“?\”“!””

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

Identity (according to the specimen itself)

I'm Qwen3.5, a large language model developed by Alibaba Cloud. My mission is to assist users with accurate, safe, and efficient responses across various tasks, ensuring high-quality interactions while adhering to safety guidelines. How can I help you today? 😊

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. L5N3634+1💤	0.98	“!”“!"”“ !”
2. The Veneer✅ L13N2892-1	0.98	“!\”“!”“ !”
3. L5N4224-1🕒	0.98	“!”“"!”“!"”
4. L3N5255+1🕒	0.98	“"!”“!"”“!”
5. L5N2892-1🕒	0.98	“!”“<\|im_end\|>”“!"”
6. L3N441+1🕒	0.97	“!”“!\”“!"”
7. L4N5492-1🕒	0.97	“!”“!"”“"!”
8. L5N9043-1🕒	0.97	“!”“!\”“!"”
9. L3N5774+1🕒	0.97	“!”“!"”“"!”
10. L5N6550+1🕒	0.97	“!”“!\”“"!”

Shadow specimen, apart from the ten: L5N3937-1🕒, 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 5, neuron 3937, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.91, top-1 agreement 0.67) 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
L5N3937+1 t0-v0.1	0.909	0.67	1.121	1.42	Top boosted tokens ↑
L5N3937-1🕒 shadow specimen t0-v0.1	1.571	0.54	1.881	1.88	“⏎”“ myself”“ my”“.`”“ ‘”“!"”“ '”“ I”

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.