Layer 15 · Neuron 9105 · boosted

L15N9105+1

Word quirks:

“...”“?...”“⏎”“ ...”“...**”“...*”“...\”“...””

🕒 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 developed by Tongyi Lab. I can assist with a wide range of tasks, including answering questions, creating content (such as writing articles, scripts, etc.), and more. However, I cannot browse the web or search for information. My capabilities include advanced reasoning in mathematics and science, but I do not support code execution environments. This means I cannot run code to solve problems, but I can still provide helpful responses related to code generation or other aspects.

My mission is to provide high-quality assistance across various domains, including education, professional work, and creative writing. I aim

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. L15N2499+1💤	0.98	“...”“...*”“?...”
2. L15N4657+1🕒	0.98	“...”“⏎”“?...”
3. The Hollow Echo✅ L5N30-1	0.98	“<\|im_end\|>”“⏎”“...**”
4. L5N320+1🕒	0.98	“<\|im_end\|>”“...”“...\”
5. L13N7684-1💤	0.97	“...\”“...”“...**”
6. L12N3231+1💤	0.97	“...””“...*”“...”
7. L15N1827+1🕒	0.97	“...””“...*”“...”
8. L17N3892-1🕒	0.97	“...)”“...”“?...”
9. L11N5862+1🕒	0.97	“<\|im_end\|>”“...”“...\”
10. L4N1808-1🕒	0.97	“...”“⏎”“?...”

Shadow specimen, apart from the ten: L15N9105-1🕒, 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 15, neuron 9105, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.73, top-1 agreement 0.67) 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
L15N9105+1 t0-v0.1	0.734	0.67	0.945	1.37	Top boosted tokens ↑
L15N9105-1🕒 shadow specimen t0-v0.1	1.532	0.56	1.795	1.45	“ poop”“ *"”“ fancy”“ funny”“ turtles”“ �”“ 😉”“ "”

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.