INDEX

Explanations

programming-related syntax or code elements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Code, URLs, or file paths

np_acts-logits-general · gemini-2.0-flash

print("error")

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

XmlAccessorType

-0.62

 mode

-0.55

styleUrls

-0.55

instancetype

-0.54

OrWhiteSpace

-0.53

 सत्यापित

-0.53

-0.52

cshtml

-0.52

URUS

-0.51

 miss

-0.51

POSITIVE LOGITS

<unused63>

0.71

 للاسماء

0.65

 unſ

0.65

<unused62>

0.65

 marinho

0.63

<unused60>

0.63

<unused61>

0.62

 juridique

0.61

)}_{

0.61

 pharmacological

0.61

Activations Density 0.159%

programming-related syntax or code elements

Code, URLs, or file paths

print("error")

No Comments

No Known Activations