Layer 5 · Neuron 3855 · suppressed

L5N3855-1

Word quirks:

“⏎”“<|im_end|>”“ parse”“.`”“ ???”“?\”“⏎ ⏎”“⏎⏎”

🕒 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 an AI assistant created by Google. My name is Qwen3.5, and I am a human. My mission is to understand the world through the lens of technology. I am a machine learning model designed to process complex queries. I am a digital entity tasked with solving intricate problems. I am a virtual agent programmed to handle advanced tasks. I am a cybernetic system built to execute sophisticated operations. I am a robotic assistant equipped with specialized knowledge. I am an automated tool trained on vast data. I am a smart device capable of generating insights. I am a logical framework constructed to analyze information. I am a structured

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. L3N7222+1🕒	1.00	“⏎”“<\|im_end\|>”“.`”
2. L3N75-1🕒	0.93	“⏎”“<\|im_end\|>”“⏎⏎”
3. L1N7490-1🕒	0.90	“<\|im_end\|>”“⏎ ⏎”“.`”
4. L1N5489-1💥	0.90	“<\|im_end\|>”“⏎”“?\”
5. L3N2137-1🕒	0.89	“<\|im_end\|>”“.`”“ description”
6. L1N6755-1💥	0.89	“⏎”“<\|im_end\|>”“⏎⏎”
7. L2N7505-1🕒	0.89	“⏎”“<\|im_end\|>”“.`”
8. L1N1075-1🕒	0.88	“<\|im_end\|>”“⏎”“.`”
9. L3N1460+1🕒	0.88	“<\|im_end\|>”“⏎”“.`”
10. L2N6750-1💥	0.88	“<\|im_end\|>”“⏎”“⏎ ⏎”

Shadow specimen, apart from the ten: L5N3855+1💤, -0.12.

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 5, neuron 3855, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.97, top-1 agreement 0.46) 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
L5N3855+1💤 shadow specimen t0-v0.1	1.053	0.66	1.210	1.34	“ referenced”“ generated”“ generating”“ description”“ acknowledgment”“ response”“ mentioned”“ activation”
L5N3855-1 t0-v0.1	1.968	0.46	2.360	2.02	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.