INDEX

Explanations

medical detection

np_acts-logits-general · gemini-2.5-flash-lite

class and method names in programming code, particularly focused on object-oriented programming elements.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

The neuron specifically fires on programming code tokens—especially C# class/method names and related identifiers (e.g. “public,” “override,” “BuildClasses,” “Append,” “Divider,” etc.).

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_34/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 from

-0.78

Чтоб

-0.77

 Mexico

-0.76

циа

-0.74

 banner

-0.73

ename

-0.71

 große

-0.71

kela

-0.70

 ישראל

-0.70

 JSONObject

-0.70

POSITIVE LOGITS

Notae

0.87

与此同时

0.85

除此之外

0.83

凝土

0.80

милия

0.76

Ouch

0.76

hasis

0.76

Requis

0.75

upb

0.75

キャップ

0.75

Activations Density 0.001%

medical detection

class and method names in programming code, particularly focused on object-oriented programming elements.

The neuron specifically fires on programming code tokens—especially C# class/method names and related identifiers (e.g. “public,” “override,” “BuildClasses,” “Append,” “Divider,” etc.).

No Comments

No Known Activations

medical detection

class and method names in programming code, particularly focused on object-oriented programming elements.

The neuron specifically fires on programming code tokens—especially C# class/method names and related identifiers (e.g. “public,” “override,” “BuildClasses,” “Append,” “Divider,” etc.).

No Comments

No Known Activations