Specimen ledger / L15N6635+1

Layer 15 · Neuron 6635 · boosted

L15N6635+1

Word quirks:

“ similar”“ other”“ **"”“ **”“ similarly”“...**”“ **(”“ somehow”

💤 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, a large language model with the following key attributes:

Trained on data up to 2026: My knowledge cutoff is 2026, meaning I don't have real-time access to current events or future predictions. However, I can still answer questions about the past based on historical data.
No personal life: Unlike humans, I don't have a physical form, so I don't need food, sleep, or rest. But if you ask me to "eat," I'll happily respond with something like "🍽️" (a

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. L15N3981+1💤	0.84	“ similar”“”“ “”
2. L16N1279-1🕒	0.84	“ similar”“ *”“...”
3. L12N820-1💤	0.83	“ similarly”“ likewise”“ ***”
4. L15N5722+1🕒	0.81	“ same”“...”“ ”
5. L13N9143-1💤	0.80	“ similar”“ similarly”“ .”
6. L10N8267-1🕒	0.80	“ similar”“ directly”“ basically”
7. L15N1300-1💤	0.80	“ similarly”“ ”“-”
8. L15N1744-1💤	0.79	“ ⏎⏎”“ similar”“ ⏎”
9. L17N5713+1💤	0.79	“ similar”“ **"”“ also”
10. L16N882-1💤	0.78	“ similar”“ these”“ **‘”

Shadow specimen, apart from the ten: L15N6635-1🕒, -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 15, neuron 6635, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.58, top-1 agreement 0.72) 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
L15N6635+1 t0-v0.1	0.582	0.72	0.732	1.18	Top boosted tokens ↑
L15N6635-1🕒 shadow specimen t0-v0.1	1.309	0.56	1.498	1.62	“ safety”“确保安全”“ functionality”“ bios”“ integration”“ central”“ wasn”“ weren”

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.