Layer 17 · Neuron 1018 · suppressed

L17N1018-1

Word quirks:

“ ‘”“ here”“ during”“ “”“ seems”“’s”“ wherever”“周遭”

💤 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 Qwen3.2, a large language model with advanced capabilities in reasoning, code generation, and multimodal analysis. My mission is to assist users in solving complex problems across various domains, including:

Reasoning & Problem-Solving: Analyzing mathematical proofs, logical puzzles, and scientific challenges.
Code Generation: Writing, debugging, and explaining code in multiple programming languages.

Multimodal Analysis: Interpreting images, charts, and documents (e.g., extracting data from diagrams).

Language Translation: Supporting 100+ languages for cross-lingual communication.

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. L17N705+1🕒	0.75	“ here”“ always”“ seems”
2. L15N6947-1🕒	0.70	“ here”“ ‘”“ elsewhere”
3. L16N1978+1💤	0.70	“ here”“ “”“ ‘”
4. L15N2824+1🕒	0.69	“ here”“ in”“ “”
5. The Flattering Echo✅ L3N9139-1,L20N7310+1	0.68	“ ‘”“?“”“ “”
6. L19N1659+1💤	0.68	“ where”“ ‘”“ “”
7. L7N1336+1💤	0.68	“ ‘”“ “”“ usually”
8. L15N7275+1💤	0.66	“ ‘”“ during”“ “”
9. L19N4017+1🕒	0.66	“ here”“ HERE”“ Here”
10. L14N1323-1💤	0.66	“ here”“ “”“ reason”

Shadow specimen, apart from the ten: L17N1018+1💤, 0.22.

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 17, neuron 1018, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.51, top-1 agreement 0.76) 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
L17N1018+1💤 shadow specimen t0-v0.1	0.605	0.73	0.742	1.09	“ ”“ ”“ ”“ ”“ being”“ later”“ bigger”“ weren”
L17N1018-1 t0-v0.1	0.505	0.76	0.514	1.20	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.