Layer 9 · Neuron 4440 · boosted

L9N4440+1

Word quirks:

“ description”“ explanation”“ ⏎”“ ⏎⏎”“ explanations”“ different”“ describing”“ descriptions”

💤 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 Qwen3.5, a large language model developed by Tongyi Lab. My mission is to provide helpful, accurate, and context-aware responses to users across different cultures and languages. I am designed to be versatile, reliable, and easy to use, without requiring any specific instructions or disclaimers.

In essence:

I am: A large language model (LLM) trained on high-quality data.
You are: The user interacting with this model.

This means that when you ask me a question, I should respond in a way that is:

Helpful and relevant to the

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. L8N5528-1🕒	0.89	“ explanation”“ explaining”“ ⏎”
2. L16N9095+1🕒	0.89	“ description”“ describing”“ descriptions”
3. L9N3599-1🕒	0.86	“ description”“ descriptions”“ measure”
4. L10N7135-1💤	0.86	“ description”“ example”“ explanation”
5. L10N7717-1🕒	0.86	“ description”“ explanation”“ describing”
6. L9N6211-1💤	0.86	“ description”“ ⏎”“ descriptions”
7. L19N2468-1💤	0.85	“ description”“ explanation”“ Description”
8. L13N8417+1🕒	0.85	“ description”“ describing”“ basically”
9. L9N206+1💤	0.85	“ description”“ actually”“ explanation”
10. L17N7916-1🕒	0.84	“ explanations”“ explanation”“ description”

Shadow specimen, apart from the ten: L9N4440-1💤, -0.02.

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 9, neuron 4440, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.93, top-1 agreement 0.68) 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
L9N4440+1 t0-v0.1	0.927	0.68	1.226	1.31	Top boosted tokens ↑
L9N4440-1💤 shadow specimen t0-v0.1	0.895	0.62	1.167	1.47	“ operation”“ "”“ "”“ your”“ operations”“ usually”“ processing”“ **【”

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.