Layer 9 · Neuron 5789 · boosted

L9N5789+1

Word quirks:

“ personally”“ now”“ thinking”“ thought”“ focus”“ individual”“ directly”“ myself”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

Hello! I am Qwen3.5. My mission is to provide you with the most advanced AI assistance possible. To ensure clarity and alignment with your goals, here is a summary of my capabilities:

Core Capabilities

Advanced Reasoning: I support complex logical reasoning tasks, including mathematical problem-solving, data analysis, and strategic planning.
Context Awareness: I maintain full context awareness to ensure all information is relevant to your specific needs.
User-Centric Design: I prioritize user experience by integrating features that directly benefit you.
Task Execution: I can execute tasks such as

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. L14N4209+1🕒	0.85	“ personal”“ self”“ personally”
2. L9N9028+1💤	0.84	“ thinking”“ memory”“ personally”
3. L10N7466+1💤	0.78	“ personally”“ personal”“ myself”
4. L8N2167+1🕒	0.78	“ myself”“ personally”“ my”
5. L18N4364-1🕒	0.75	“ individually”“ individual”“ personally”
6. L6N8833+1🕒	0.75	“ privacy”“ personal”“ myself”
7. L5N3064-1🕒	0.75	“ myself”“ yourself”“ ‘”
8. L9N920-1🕒	0.74	“ my”“ thinking”“ myself”
9. L14N8150-1🕒	0.74	“ thinking”“ topics”“ info”
10. L11N1545+1💤	0.74	“ personally”“ personal”“ himself”

Shadow specimen, apart from the ten: L9N5789-1🕒, 0.01.

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 5789, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.87, 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
L9N5789+1 t0-v0.1	0.867	0.63	1.099	1.35	Top boosted tokens ↑
L9N5789-1🕒 shadow specimen t0-v0.1	2.185	0.36	2.637	2.15	“ "”“ "”“ ____”“ basically”“ seems”“ .”“ apparently”“ seemed”

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.