Layer 13 · Neuron 2828 · boosted

L13N2828+1

Word quirks:

“ minimal”“ small”“ tiny”“ smaller”“ distant”“ simpler”“ basic”“ simple”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen3.5, a large language model with a knowledge cutoff date of 2026. I can assist with tasks like summarizing information, answering questions about distant topics, or helping with coding challenges. If you need something simple, I'll do my best to provide a concise response. Would you like to explore a specific topic together? 🌟

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. L13N9083-1🕒	0.93	“ small”“ simple”“ Small”
2. L14N5489+1🕒	0.92	“ basic”“ small”“ minimal”
3. L14N8421-1🕒	0.91	“ small”“ basic”“ smaller”
4. L17N7927+1🕒	0.91	“ small”“ Small”“ simple”
5. L14N3976+1🕒	0.89	“ small”“ Small”“ smaller”
6. L14N7396+1🕒	0.89	“ small”“ minor”“ basic”
7. L13N6698+1🕒	0.88	“ tiny”“ small”“ smaller”
8. L14N7302-1🕒	0.88	“ small”“ tiny”“ basic”
9. L17N6880-1🕒	0.87	“ minimal”“ basic”“ mini”
10. L17N4793-1🕒	0.87	“ minimal”“ minor”“ smaller”

Shadow specimen, apart from the ten: L13N2828-1🕒, 0.10.

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 13, neuron 2828, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.64) 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
L13N2828+1 t0-v0.1	0.890	0.64	1.228	1.31	Top boosted tokens ↑
L13N2828-1🕒 shadow specimen t0-v0.1	1.788	0.52	2.081	1.84	“ typical”“ typically”“.:”“.\"”“:"”“:”“ perfect”“.:*”

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.