Layer 13 · Neuron 8047 · suppressed

L13N8047-1

Word quirks:

“ recently”“ now”“<|im_end|>”“ carefully”“ description”“ updated”“ recent”“ modern”

💤 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 large language model developed by Alibaba Cloud's Tongyi Lab. My mission is to redefine how I interact with users by integrating advanced capabilities into my architecture. I focus on enhancing my reasoning abilities, improving my training data, and refining my mission to ensure that I can handle complex tasks with greater precision, efficiency, and adaptability.

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. Phantom Changelog✅ L11N7045+1	0.83	“ recent”“ recently”“ now”
2. L12N7289-1💤	0.83	“ recently”“ specific”“ now”
3. L10N8526-1🕒	0.82	“ recently”“ now”“ similar”
4. L13N6019-1💤	0.82	“ new”“ now”“ recently”
5. L14N8655-1🕒	0.82	“ newer”“ updated”“ recently”
6. L21N6168+1💤	0.81	“ recent”“ Recent”“ current”
7. L22N2403+1💤	0.81	“ recent”“ recently”“ current”
8. L22N3606-1🕒	0.81	“ recent”“ current”“ Recent”
9. L9N2253-1💤	0.81	“ recently”“ lately”“ recent”
10. L1N3472-1💥	0.81	“ lately”“<\|im_end\|>”“ recently”

Shadow specimen, apart from the ten: L13N8047+1🕒, 0.11.

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 8047, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.79, top-1 agreement 0.68) 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
L13N8047+1🕒 shadow specimen t0-v0.1	1.086	0.56	1.331	1.46	“ ⏎”“<\|im_end\|>”“ additional”“⏎”“ ⏎⏎”“\n”“⏎ ⏎”“ ⏎ ⏎”
L13N8047-1 t0-v0.1	0.789	0.68	1.020	1.34	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.