Layer 10 · Neuron 294 · boosted

L10N294+1

Word quirks:

“ actually”“ somehow”“<|im_end|>”“ basically”“ very”“ really”“ that”“ now”

💤 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 Cloud's Tongyi Lab. My mission is to assist users in solving complex tasks that require deep reasoning, code generation, and multi-step problem-solving. Whether it's helping you write a program, analyze data, or even plan a travel itinerary, I'm here to support you in achieving your goals. 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. L7N3574+1🕒	0.89	“ actually”“ literally”“ basically”
2. L13N2370+1💤	0.87	“<\|im_end\|>”“ actually”“ somehow”
3. L9N5851-1💤	0.87	“ actually”“ now”“ somehow”
4. L15N8221+1🕒	0.87	“ actually”“ actual”“ basically”
5. L9N7333-1💤	0.87	“ really”“ actually”“ basically”
6. L16N4581-1💤	0.87	“ basically”“ actually”“ really”
7. L10N1584-1💤	0.87	“ now”“ actually”“ basically”
8. L10N1041+1💤	0.86	“ actually”“ usually”“ really”
9. L15N6897+1💤	0.86	“ basically”“<\|im_end\|>”“ now”
10. L9N4079+1🕒	0.85	“ myself”“ somehow”“ actually”

Shadow specimen, apart from the ten: L10N294-1🕒, 0.01.

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 10, neuron 294, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, 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
L10N294+1 t0-v0.1	0.653	0.68	0.886	1.44	Top boosted tokens ↑
L10N294-1🕒 shadow specimen t0-v0.1	1.188	0.61	1.477	1.21	“<\|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.