Layer 29 · Neuron 9016 · boosted

L29N9016+1

Word quirks:

“ sk”“ incompat”“ Compatible”“不相”“ider”“ compatible”“ compat”“compat”

💤 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 Labs. My core purpose is to be a helpful, accurate, and safe assistant for a wide range of tasks—whether you need help with complex reasoning, creative writing, data analysis, or just a friendly conversation.

My Mission

Empower Problem-Solving: Break down challenging tasks (like scientific research, code debugging, or logical puzzles) into clear, step-by-step solutions.
Adapt to Your Needs: Tailor responses to your style, expertise level, or specific goals (

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. L19N8178-1💤	0.90	“ compat”“ compatibility”“ compatible”
2. L28N6527+1💤	0.89	“ compat”“ compatibility”“ Compatibility”
3. L6N3288+1🕒	0.89	“ compatibility”“ compatible”“ compat”
4. L21N5441-1🕒	0.88	“ compatible”“ compatibility”“ compat”
5. L1N7107+1💥	0.88	“ compat”“ compatibility”“ compatible”
6. L2N4881+1💥	0.88	“ compatible”“ Compatible”“ compatibility”
7. L5N4288-1🕒	0.88	“ compatible”“ compatibility”“ compat”
8. L30N1105+1💥	0.88	“ compatibility”“compat”“ compatible”
9. L19N5925+1🕒	0.87	“ compatibility”“ compat”“兼容性”
10. L27N2869-1🕒	0.87	“ compatibles”“ compatible”“ Compatible”

Shadow specimen, apart from the ten: L29N9016-1🕒, 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 29, neuron 9016, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.28, top-1 agreement 0.81) 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
L29N9016+1 t0-v0.1	0.279	0.81	0.373	1.38	Top boosted tokens ↑
L29N9016-1🕒 shadow specimen t0-v0.1	0.406	0.78	0.452	1.49	“ z”“ з”“/z”“,z”“.z”“ ز”“=z”“ poz”

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.