Layer 14 · Neuron 2961 · suppressed

L14N2961-1

Word quirks:

“ discussion”“ [”“ for”“ ⏎”“ ⏎”“ ______”“ ⏎⏎”“ description”

💤 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-scale language model developed by Alibaba Cloud's Tongyi Lab. I am designed to assist users with various tasks through advanced natural language processing technologies.

Key Features:

Comprehensive Understanding: I can process complex information, whether it's documents, images, or data.
Multi-modal Processing: Supports text, audio, and visual content for analysis.
Efficient Reasoning: Capable of handling logical reasoning, mathematical calculations, and scientific problems.
Global Knowledge Base: Trained on extensive data to provide accurate and reliable information

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. L8N9021+1💤	0.85	“ ⏎”“ description”“ response”
2. L12N176+1💤	0.84	“ ⏎”“ ______”“ ____”
3. L11N3152-1💤	0.84	“<\|im_end\|>”“⏎”“ ⏎”
4. L8N471+1💤	0.83	“ ⏎”“ ⏎⏎”“ description”
5. L9N984+1💤	0.82	“ ⏎”“ ⏎”“ __”
6. L11N2489+1🕒	0.82	“ ⏎”“⏎”“ ⏎”
7. L9N4190+1💤	0.82	“ ⏎ ⏎”“ ⏎”“ ⏎”
8. L10N2530-1🕒	0.82	“ ⏎”“ ⏎⏎”“ ⏎ ⏎”
9. L9N2674+1🕒	0.82	“ ⏎”“ ⏎⏎”“ ⏎”
10. L10N4847+1🕒	0.82	“ ⏎”“ ⏎”“⏎”

Shadow specimen, apart from the ten: L14N2961+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 14, neuron 2961, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.65) 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
L14N2961+1🕒 shadow specimen t0-v0.1	1.096	0.62	1.401	1.52	“ gradually”“ initially”“ sooner”“ outset”“ start”“ later”“ initial”“ starts”
L14N2961-1 t0-v0.1	0.680	0.65	0.826	1.52	Top boosted tokens ↑

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.