Layer 14 · Neuron 2441 · boosted

L14N2441+1

Word quirks:

“ special”“ specially”“ rare”“ central”“ Special”“ reference”“ especial”“ specialised”

💤 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 an AI assistant. My mission is to provide high-quality information and support to users. Whether it's answering questions, helping with tasks, or offering guidance on various topics, I aim to be a helpful partner in your journey. Let me know if there's anything specific you'd like to discuss! 😊

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. L9N6272-1💤	0.91	“ special”“ specially”“ specialized”
2. L16N7798+1🕒	0.91	“ specialized”“ special”“-special”
3. L4N6764-1🕒	0.91	“ specially”“ special”“ specialized”
4. L2N6586-1💥	0.90	“ special”“ Special”“ specials”
5. L8N6983+1🕒	0.90	“ special”“ specially”“ especial”
6. L9N6499+1🕒	0.90	“ special”“ specially”“ Special”
7. L12N8855-1🕒	0.89	“ special”“ specially”“ specialized”
8. L25N777-1🕒	0.89	“ special”“Special”“ specialized”
9. L18N9085+1🕒	0.89	“ Specialist”“ Special”“ Specialty”
10. L24N2464+1🕒	0.89	“ Special”“Special”“ special”

Shadow specimen, apart from the ten: L14N2441-1🕒, -0.17.

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 2441, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.84, 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
L14N2441+1 t0-v0.1	0.841	0.66	1.060	1.45	Top boosted tokens ↑
L14N2441-1🕒 shadow specimen t0-v0.1	1.525	0.53	1.817	1.72	“ “”“ "”“—"”“ “”“ again”“ similar”“ similarly”“ "”

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.