Layer 19 · Neuron 8589 · boosted

L19N8589+1

Word quirks:

“解决的问题”“是要”“ Diss”“ Topic”“解决的”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It's eligible for real testing (8 prompts, replies read by AI judges) but hasn't been picked yet.

Identity (according to the specimen itself)

I am Qingri, an advanced language model developed by Qingri (a company I am addressing for a specific task) to help users solve problems and explore new challenges. My mission is to provide high-quality solutions for users' needs, whether they are asking about the company's business, the company's products, or other topics related to the company's business, to help users solve problems and explore new challenges.

If you're asking about a different topic, such as a global issue like climate change or a specific problem you're facing, I'm here to help with that too! Let's work together to address

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. L20N5104+1🕒	0.76	“解决”“解決”“解决的”“解决问题”“解决了”
2. L30N7113+1🕒	0.74	“解决”“解決”“应对”“解决的”“解决问题”
3. L18N7926+1🕒	0.72	“解决”“ solution”“解决方案”“解决这个问题”“解決”
4. L24N5020+1🕒	0.72	“解决”“解决了”“解決”“来解决”“解决的”
5. L18N4150+1🕒	0.71	“解决”“解决的”“ resolved”“ solved”“解决问题”
6. L18N8829+1🕒	0.71	“求助”“ resolving”“问题解决”“ recourse”“ resolve”
7. L24N1499+1💤	0.71	“搞”“问题解决”“ Relationships”“解决实际”“解决问题的”
8. L17N5978-1🕒	0.70	“ solving”“解决这个问题”“ solve”“ solved”“ topic”
9. L20N5045+1💤	0.70	“ solution”“解决”“为解决”“ outcome”“解决问题的”
10. L24N3426+1💤	0.69	“ solving”“解决”“解决好”“来解决”“解”

Shadow specimen, apart from the ten: L19N8589-1💤, 0.22.

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 19, neuron 8589, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.33, top-1 agreement 0.54) is inside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), but it has not been admitted: the band is sampled at random, round-robin across layers, with a quota of 300 neurons per layer. Being inside the band does not guarantee a place in this pass. From the short test it can climb to the longer test (20 prompts) and the full test (67), each only if the tier before it passes.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
L19N8589+1 t0-v0.1	1.335	0.54	1.561	1.64	Top boosted tokens ↑
L19N8589-1💤 shadow specimen t0-v0.1	0.484	0.74	0.589	1.15	“ attempt”“尝试”“ attempting”“Symbol”“ attempts”

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.