Layer 13 · Neuron 360 · boosted

L13N360+1

Word quirks:

“ info”“ usual”“ relevant”“ docs”“ pertinent”“ aligned”“ presently”“ usually”

🕒 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)

Hello! I am Qwen3.5, the latest version of Qwen. My mission is to provide high-quality assistance across all domains.

---

🌐 Mission Scope

Core Competencies:
Text Generation: Create coherent content based on user intent.
Multilingual Support: Fluent in over 100 languages worldwide.
Visual Analysis: Deep understanding of complex diagrams and charts.
Code & Math: Solve math problems with high accuracy.
Agent Planning: Execute multi-step tasks autonomously.
**

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. L9N966+1🕒	0.82	“ info”“ queries”“ relevant”
2. L7N9107+1🕒	0.80	“ related”“ \"”“ info”
3. L9N5549+1🕒	0.79	“ usual”“ related”“ relevant”
4. L8N2801+1💤	0.78	“⏎”“ relevant”“<\|im_end\|>”
5. L16N3769+1🕒	0.76	“ relevant”“ typical”“ referenced”
6. L10N8405+1💤	0.76	“ related”“ usually”“!:”
7. L12N3145-1💤	0.76	“ info”“ actually”“ relevant”
8. L7N5569-1💤	0.75	“ thinking”“ usually”“ usual”
9. L12N4104-1💤	0.75	“ based”“ now”“ corresponding”
10. L12N4327-1💤	0.75	“ overall”“ now”“ relevant”

Shadow specimen, apart from the ten: L13N360-1🕒, -0.09.

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 360, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.59) 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
L13N360+1 t0-v0.1	0.888	0.59	1.158	1.62	Top boosted tokens ↑
L13N360-1🕒 shadow specimen t0-v0.1	0.868	0.67	1.123	1.70	“ ⏎”“⏎”“<\|im_end\|>”“⏎ ⏎”“ ____”“ ⏎⏎”“ ⏎ ⏎”“ maybe”

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.