Layer 16 · Neuron 4884 · boosted

L16N4884+1

Word quirks:

“ anyway”“ nonetheless”“ actually”“ anyways”“ however”“ yet”“ nevertheless”“ thus”

💤 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 Alibaba Tongyi Lab. My core capabilities include advanced reasoning, enhanced language understanding, and the ability to handle complex tasks across domains. I can assist users with writing, coding, data analysis, and more.

My mission is to empower users to achieve their goals through efficient problem-solving. This includes:

Assisting users in completing tasks (e.g., drafting documents, debugging code, analyzing datasets).
Providing high-quality responses to queries (e.g., answering questions about topics like science, history, or law).
**Support

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. L27N576+1🕒	0.90	“ anyways”“ anyway”“ nevertheless”
2. L16N3621-1💤	0.89	“ moreover”“⏎”“ anyway”
3. L8N8050-1💥	0.89	“ anyways”“ thus”“ anyway”
4. L30N6320+1💥	0.88	“ anyway”“ anyways”“ nonetheless”
5. L17N4650-1🕒	0.88	“ anyway”“ rather”“ actually”
6. L19N6751+1🕒	0.87	“ nevertheless”“ nonetheless”“ however”
7. L2N1724-1💥	0.87	“ nevertheless”“ however”“ anyway”
8. L17N527-1💤	0.87	“ moreover”“ hence”“ accordingly”
9. L17N4302-1🕒	0.87	“ though”“ however”“ nonetheless”
10. The Retry Loop✅ L16N3423-1	0.87	“ lengths”“ anyway”“ nonetheless”

Shadow specimen, apart from the ten: L16N4884-1🕒, 0.09.

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 16, neuron 4884, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.60, top-1 agreement 0.70) 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
L16N4884+1 t0-v0.1	0.595	0.70	0.724	1.50	Top boosted tokens ↑
L16N4884-1🕒 shadow specimen t0-v0.1	0.916	0.62	1.281	1.55	“ .”“-”“ .**”“?:”“?-”“-n”“-separated”“l”

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.