Layer 15 · Neuron 6893 · suppressed

L15N6893-1

Word quirks:

“ your”“ broader”“ you”“ further”“ my”“ deeper”“ continuation”“ Your”

💤 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 Lab. My capabilities span a wide range of tasks, from creative writing to complex reasoning, with a focus on accuracy and efficiency.

My mission is to provide high-quality assistance that aligns with your needs, whether you're looking for detailed analysis, creative content, or technical solutions. I aim to be a reliable partner in understanding, learning, and collaborating with you—ensuring clarity, precision, and relevance in every interaction.

Let me know if you'd like to explore my capabilities further! 😊

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. L13N425+1🕒	0.90	“ your”“ broader”“ you”
2. The Leaner✅ L12N4531-1	0.85	“ your”“ closer”“ more”
3. L16N1770-1💤	0.80	“ you”“ for”“ more”
4. L15N2947-1💤	0.80	“ now”“ your”“ yours”
5. L15N8894-1💤	0.78	“ further”“ already”“ other”
6. L13N6213-1🕒	0.78	“ your”“ clearer”“ Your”
7. L15N2362+1💤	0.78	“ you”“ YOU”“ further”
8. L13N7464+1🕒	0.77	“ your”“ thus”“ interesting”
9. L11N4648-1💤	0.77	“ your”“ similar”“ always”
10. L14N8796+1💤	0.77	“ yourself”“ continued”“ continuing”

Shadow specimen, apart from the ten: L15N6893+1🕒, 0.21.

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 15, neuron 6893, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.50, top-1 agreement 0.72) 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
L15N6893+1🕒 shadow specimen t0-v0.1	1.967	0.39	2.281	1.89	“ allowed”“ accordingly”“ approach”“ __”“ corresponding”“ allow”“ treatment”“ solution”
L15N6893-1 t0-v0.1	0.496	0.72	0.692	1.12	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.