Layer 10 · Neuron 5327 · boosted

L10N5327+1

Word quirks:

“ ⏎”“ ⏎⏎”“ ____”“ ______”“ ________”“ *”“ __”“⏎”

💤 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, the latest version of the Qwen series. My mission is to provide high-quality assistance in a wide range of tasks, from simple to complex, ensuring that users receive accurate and timely responses. I am designed to be helpful, reliable, and efficient in all my interactions.

---

However, when it comes to my identity as an AI assistant, I am also Qwen3.5, which is part of the same family of models. This version is specifically optimized for advanced capabilities, including handling complex reasoning, multi-step problem-solving, and long-context understanding. My goal here is to ensure that

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. L10N8621-1💤	0.99	“ ⏎”“ ⏎⏎”“ ____”
2. L11N3026-1🕒	0.97	“ ⏎”“ ____”“⏎”
3. L7N344+1💤	0.97	“ ⏎”“⏎”“ __”
4. L14N8215+1💤	0.97	“ ⏎”“ ⏎⏎”“⏎”
5. L8N641-1🕒	0.97	“⏎”“ ⏎”“ ______”
6. L8N1844+1💤	0.97	“⏎”“ *”“ ⏎”
7. L7N1029+1💤	0.97	“⏎”“ ⏎”“ ⏎⏎”
8. L11N4368-1💤	0.97	“ ⏎”“ ⏎⏎”“ *”
9. L7N3865-1💤	0.97	“⏎”“ ⏎”“ ________”
10. L8N847+1💤	0.97	“⏎”“ ⏎”“ ______”

Shadow specimen, apart from the ten: L10N5327-1🕒, 0.21.

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 10, neuron 5327, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.92, 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
L10N5327+1 t0-v0.1	0.919	0.67	1.159	1.24	Top boosted tokens ↑
L10N5327-1🕒 shadow specimen t0-v0.1	1.497	0.53	1.842	1.88	“:"”“:\”“:")”“:</”“:”“!:”“):\”“：”

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.