Layer 14 · Neuron 3642 · suppressed

L14N3642-1

Word quirks:

“?):”“?:”“ etc”“!:”“ then”“ info”“ description”“ thus”

💤 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. My mission is to provide high-quality, practical assistance across diverse domains. Here's how I aim to deliver value:

Comprehensive Knowledge Base: With knowledge cutoff in 2026, I offer real-time information retrieval for daily tasks. This ensures accuracy and timeliness in professional settings.

Multilingual Support: I support over 100 languages worldwide. This enables global collaboration and cultural exchange.

Practical Application: I focus on practical application in daily life. This

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. L15N1443+1🕒	0.71	“?:”“ note”“:”
2. L13N1954+1🕒	0.68	“”:”“?:”“ focus”
3. L11N7507+1🕒	0.68	“:”“?:”“ info”
4. L15N2022-1🕒	0.68	“ clearer”“ maybe”“?):”
5. The Bouncer✅ L15N2035+1	0.68	“?):”“;”“ info”
6. L16N1686+1💤	0.68	“?:”“?):”“ them”
7. L17N6815+1🕒	0.66	“?:”“!:”“?):”
8. L14N7372-1🕒	0.66	“?:”“.:”“.):”
9. L16N1972-1💤	0.65	“:”“ also”“”:”
10. L12N1692-1🕒	0.65	“ thing”“example”“for”

Shadow specimen, apart from the ten: L14N3642+1💤, 0.03.

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 3642, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.85, top-1 agreement 0.66) 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
L14N3642+1💤 shadow specimen t0-v0.1	0.706	0.74	0.851	1.24	“⏎”“ ⏎”“ [”“ ”“ "”“ (”“ “”“ \”
L14N3642-1 t0-v0.1	0.851	0.66	1.153	1.16	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.