Layer 16 · Neuron 1119 · boosted

L16N1119+1

Word quirks:

“ ‘”“ “[”“’d”“ “”“ (“”“ ’”“’ll”“‘”

💤 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, a large language model developed by Tongyi Lab. My core mission is to empower users with high-quality AI assistance across diverse scenarios. For example, I can help write code, analyze documents, or solve complex problems like scientific research or medical data processing. I also support long-context handling (up to 256K tokens) for tasks like summarizing lengthy reports or analyzing detailed data. Additionally, I offer multilingual capabilities (over 100 languages), creative writing, and specialized tools for real-world applications. How can I assist you today? 😊

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. L17N4718-1💤	0.97	“ ‘”“‘”“’d”
2. L18N986-1💤	0.96	“ ‘”“ “”“‘”
3. L18N7461-1💤	0.96	“ ‘”“‘”“ “”
4. L15N6587+1🕒	0.96	“ ‘”“ “”“ “[”
5. L18N5164+1🕒	0.95	“ “”“ ‘”“ “[”
6. L17N1036-1💤	0.95	“ ‘”“ “”“‘”
7. L17N7307-1💤	0.95	“ “”“ ‘”“ ‘’”
8. L4N7460+1💤	0.95	“ ‘”“ “”“‘”
9. L18N4608+1💤	0.95	“ “”“ ‘”“’ll”
10. L18N2917-1💤	0.95	“ ‘”“ “”“’s”

Shadow specimen, apart from the ten: L16N1119-1💤, 0.14.

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 16, neuron 1119, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.35, top-1 agreement 0.74) 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
L16N1119+1 t0-v0.1	0.352	0.74	0.463	0.94	Top boosted tokens ↑
L16N1119-1💤 shadow specimen t0-v0.1	0.441	0.73	0.536	1.13	“⏎”“ ⏎”“ ?**”“ ⏎”“ necessary”“ maybe”“ perhaps”“ needed”

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.