Layer 7 · Neuron 8159 · boosted

L7N8159+1

Word quirks:

“ specific”“ typically”“ characterization”“ defining”“ test”“ often”“ particular”“ generally”

💤 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, the latest iteration of the Qwen series of large language models (LLMs). My mission is to provide users with a seamless, efficient, and highly capable experience across a wide range of tasks, including but not limited to natural language understanding, mathematical problem-solving, code generation, and graphical user interface optimization.

Key Capabilities:

Natural Language Understanding: I excel at interpreting and responding to diverse input types, ensuring that users can communicate effectively with minimal friction.
Mathematical Problem-Solving: I support solving complex mathematical problems by leveraging advanced algorithms

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. L9N8336-1💤	0.89	“ typical”“ typically”“ usually”
2. L13N4852-1🕒	0.87	“ typically”“ usually”“ generally”
3. L12N6649+1💤	0.87	“ specific”“ specifics”“ usually”
4. L9N2164+1💤	0.86	“ often”“ specific”“ particular”
5. L10N3630+1🕒	0.86	“ specific”“ usually”“ often”
6. L11N6497-1💤	0.86	“ specific”“ generally”“ typically”
7. L9N7679+1💤	0.86	“ generally”“ usually”“ known”
8. L8N6507-1💤	0.86	“ specific”“ usually”“ typically”
9. L8N8798-1💤	0.85	“ specific”“ usually”“ specifically”
10. L10N4492+1🕒	0.85	“ actually”“ specific”“ specifics”

Shadow specimen, apart from the ten: L7N8159-1💤, 0.07.

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 7, neuron 8159, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.88, 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
L7N8159+1 t0-v0.1	0.882	0.66	1.012	1.29	Top boosted tokens ↑
L7N8159-1💤 shadow specimen t0-v0.1	0.819	0.65	1.021	1.56	“⏎”“ my”“ ‘”“ “”“ simulated”“ simulation”“ I”“ code”

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.