Layer 12 · Neuron 2595 · boosted

L12N2595+1

Word quirks:

“ ‘”“⏎”“...””“<|im_end|>”“ ______”“ ⏎”“...\”“ ____”

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

Identity (according to the specimen itself)

Hello! I am Qwen3.5, the latest large language model developed by Tongyi Lab. My mission is to empower users with advanced capabilities in reasoning, coding, and multi-modal understanding. How can I assist 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. L6N9009+1💤	1.00	“⏎”“<\|im_end\|>”“ ____”
2. L6N6617-1🕒	0.98	“<\|im_end\|>”“ ⏎”“ ______”
3. L7N3456-1💤	0.96	“<\|im_end\|>”“⏎”“ ____”
4. L6N3171-1💤	0.96	“<\|im_end\|>”“ ‘”“ _____”
5. L5N1819-1💤	0.96	“⏎”“<\|im_end\|>”“\n”
6. L4N820-1🕒	0.96	“⏎”“ ⏎”“ ‘”
7. L6N8399-1💤	0.96	“<\|im_end\|>”“⏎”“ ‘”
8. L2N8757+1💥	0.96	“<\|im_end\|>”“⏎”“ ⏎”
9. L5N160+1🕒	0.96	“⏎”“ ‘”“<\|im_end\|>”
10. L5N1348+1🕒	0.95	“<\|im_end\|>”“⏎”“?\”

Shadow specimen, apart from the ten: L12N2595-1🕒, -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 12, neuron 2595, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.56, 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
L12N2595+1 t0-v0.1	0.557	0.68	0.727	1.24	Top boosted tokens ↑
L12N2595-1🕒 shadow specimen t0-v0.1	1.117	0.62	1.275	1.04	“ --”“ info”“ always”“ doesn”“ +”“ somehow”“ combo”“ combos”

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.