INDEX

Explanations

numerical codes or identifiers

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

photo filenames with numbers

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

tagHelperRunner

-0.69

 Мексичка

-0.65

KommentareTeilen

-0.62

TagMode

-0.61

Diweddarwch

-0.60

CppMethod

-0.56

 صوتيه

-0.55

httphttps

-0.53

ghijklmnop

-0.53

 '\\;'

-0.53

POSITIVE LOGITS

 photographed

0.62

 photographing

0.57

 photographs

0.57

 photograph

0.56

 photo

0.53

 Photographic

0.53

photo

0.53

 photography

0.52

 photographer

0.52

photography

0.52

Activations Density 0.008%

numerical codes or identifiers

photo filenames with numbers

No Comments

No Known Activations