Layer 11 · Neuron 1029 · suppressed

L11N1029-1

Word quirks:

“ similar”“ different”“ sometimes”“ every”“ many”“ often”“ just”“ directly”

🕒 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, the latest large language model developed by Alibaba Cloud. I am part of a series of models designed to help users with various tasks, from answering questions to solving complex problems.

What Makes Me Different?

While I share many capabilities with other top models like GPT-4o or Claude 3, I have been specifically trained on a massive dataset that includes:

Mathematical reasoning (solving equations, logic puzzles).
Scientific principles (understanding physics, chemistry, biology).
Real-world applications

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. L14N5615+1🕒	0.90	“ sometimes”“ often”“ many”
2. L12N5240-1🕒	0.85	“ similar”“ often”“ sometimes”
3. L13N6565+1🕒	0.85	“ different”“ various”“ similar”
4. L13N2362-1💤	0.85	“ often”“ similar”“ sometimes”
5. L13N280-1💤	0.84	“ sometimes”“ often”“ usually”
6. L15N5842-1💤	0.83	“ sometimes”“ similar”“ even”
7. L13N5523+1🕒	0.83	“ different”“ sometimes”“ specific”
8. L16N3768-1🕒	0.83	“ often”“ sometimes”“ different”
9. L12N5479-1💤	0.82	“ usually”“ similar”“ basically”
10. L10N7053-1🕒	0.82	“ certain”“ particular”“ sometimes”

Shadow specimen, apart from the ten: L11N1029+1🕒, 0.04.

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 11, neuron 1029, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.81, top-1 agreement 0.62) 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
L11N1029+1🕒 shadow specimen t0-v0.1	1.136	0.61	1.434	1.37	“ ⏎”“ ‘’”“ ⏎ ⏎”“ \"”“ nowadays”“ info”“ ””“ nicely”
L11N1029-1 t0-v0.1	0.807	0.62	0.964	1.30	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.