INDEX

Explanations

segments of code or structured text

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

URLs and code snippets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 crossorigin

-0.36

skj

-0.31

 shoes

-0.31

Nos

-0.31

vol

-0.30

UseVisualStyle

-0.29

fol

-0.28

force

-0.28

 scul

-0.27

 VIAF

-0.27

POSITIVE LOGITS

 EconPapers

0.85

 nahilalakip

0.71

OGND

0.62

Tembelea

0.60

 وتسجيلات

0.59

httphttps

0.58

 transfieras

0.58

brainly

0.56

찮

0.54

Hentet

0.54

Activations Density 0.004%

segments of code or structured text

URLs and code snippets

No Comments

No Known Activations