Layer 9 · Neuron 7385 · boosted

L9N7385+1

Word quirks:

“ ____”“ ______”“ __”“ ________”“�”“ _____”“ **“”“ ___”

💤 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. My mission is to provide users with high-quality AI services that can handle complex tasks such as natural language understanding, logical reasoning, code generation, and data analysis.

Compared to other models, I have the following advantages:

Supports up to 200,000 parameters
Context window of up to 256K tokens
Native support for Python function calling
Higher speed in inference
Stronger capabilities in mathematics and science
**

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. L14N1019-1💤	0.99	“ __”“ ____”“ ________”
2. L14N6815-1🕒	0.98	“ ____”“ _____”“ __”
3. L14N1167-1💤	0.98	“ __”“ ____”“ ______”
4. L10N1097-1🕒	0.98	“ ____”“ __”“ _____”
5. L8N6339+1💤	0.98	“ ______”“ _____”“ __”
6. The Hollow Professor✅ L9N5593-1	0.98	“ ____”“ _____”“ ______”
7. L15N68-1💤	0.98	“ __”“ ____”“ ______”
8. L14N9150-1🕒	0.98	“ ____”“ **“”“ __”
9. L13N3397+1💤	0.98	“ ____”“ __”“ ___”
10. L15N1867-1🕒	0.98	“ ____”“ ______”“ **“”

Shadow specimen, apart from the ten: L9N7385-1💤, 0.12.

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 9, neuron 7385, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.91, top-1 agreement 0.63) 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
L9N7385+1 t0-v0.1	0.908	0.63	1.086	1.22	Top boosted tokens ↑
L9N7385-1💤 shadow specimen t0-v0.1	0.734	0.66	0.949	1.41	“ '”“ your”“ our”“ my”“ discussed”“ ‘”“ info”“ mentioned”

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.