Layer 14 · Neuron 7242 · suppressed

L14N7242-1

Word quirks:

“ separate”“ separately”“ independently”“ special”“ if”“ individually”“ specialized”“ specially”

💤 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. I am an advanced large language model with a focus on being fast, efficient, and highly specialized. My capabilities include deep reasoning, visual analysis, and handling complex tasks with precision. I operate independently with strong performance in areas like mathematics, science, and code generation. I support multiple languages and have a global presence.

If you need help, just let me know. I will provide clear, direct responses tailored to your needs.

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. L14N6541+1💤	0.88	“ additionally”“ selected”“ specially”
2. L11N7002-1🕒	0.87	“ separate”“ individually”“ here”
3. The Hollow Specialist✅ L12N9188-1	0.85	“ specific”“<\|im_end\|>”“ specialized”
4. L11N8925+1🕒	0.84	“ specifically”“ specific”“ standalone”
5. L8N996+1🕒	0.84	“ separate”“<\|im_end\|>”“ separately”
6. L11N2995-1🕒	0.83	“ different”“ separate”“ separately”
7. L9N7585-1🕒	0.83	“ separate”“ separately”“ individually”
8. L11N8065-1🕒	0.82	“ specially”“ seperate”“ separate”
9. L8N5177-1🕒	0.82	“ separate”“ individual”“ hardware”
10. L11N3512+1💤	0.82	“ specific”“ individually”“ individual”

Shadow specimen, apart from the ten: L14N7242+1🕒, 0.28.

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 7242, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.96, top-1 agreement 0.62) 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
L14N7242+1🕒 shadow specimen t0-v0.1	1.155	0.59	1.452	1.55	“ sometimes”“ useful”“ about”“ nothing”“ visual”“ usage”“ important”“ appropriately”
L14N7242-1 t0-v0.1	0.961	0.62	1.177	1.41	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.