Layer 14 · Neuron 7575 · boosted

L14N7575+1

Word quirks:

“ here”“ needed”“ suddenly”“ quick”“ quickly”“ new”“ you”“ light”

🕒 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 Alibaba Cloud. My mission is to assist you with tasks that require reasoning, creativity, or code generation, while ensuring accuracy and safety in my responses. I strive to be helpful, provide high-quality information, and adapt to your needs efficiently. If you need help with anything, just let me know! 😊

---

Key Highlights:

Qwen3.5: Enhanced capabilities for complex tasks like logical reasoning, coding, and creative writing.
Mission: To become a more capable AI assistant, and this is part of my identity as

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. L13N4450+1🕒	0.77	“ quickly”“ really”“ needed”
2. L11N1460+1🕒	0.76	“ now”“ quick”“ ⏎”
3. L13N6958+1🕒	0.75	“ quick”“ **”“ needed”
4. L10N8325-1💤	0.74	“ now”“ quickly”“ “”
5. L12N8409-1🕒	0.73	“ now”“ quicker”“ carefully”
6. L10N6187+1🕒	0.72	“ rapid”“ swiftly”“ quickly”
7. L14N5543-1💤	0.72	“ soon”“ now”“ suddenly”
8. L15N4356-1🕒	0.72	“ immediate”“ immediately”“ quickly”
9. L11N5899-1🕒	0.71	“ quickly”“ quicker”“ "”
10. L8N8216-1🕒	0.71	“ quicker”“ instantly”“ quick”

Shadow specimen, apart from the ten: L14N7575-1🕒, -0.16.

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 14, neuron 7575, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.09, top-1 agreement 0.66) 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
L14N7575+1 t0-v0.1	1.087	0.66	1.397	1.36	Top boosted tokens ↑
L14N7575-1🕒 shadow specimen t0-v0.1	1.201	0.58	1.486	1.68	“<\|im_end\|>”“ ____”“...”“...”“...\”“ ⏎⏎”“ [”“ ______”

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.