INDEX

Explanations

references to people

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

people doing or being

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

 الرياضيه

-0.36

Myself

-0.34

(@"%@",

-0.34

 Myself

-0.32

 saumon

-0.32

])):

-0.30

 अलावा

-0.30

 समीक्षाएं

-0.30

 يتيمه

-0.29

','',

-0.29

POSITIVE LOGITS

SequentialGroup

0.82

AnchorStyles

0.63

tagHelperRunner

0.59

पया

0.57

 تضيفلها

0.56

everybody

0.55

encils

0.55

wiliwch

0.53

ésultats

0.52

earcher

0.52

Activations Density 0.027%

references to people

people doing or being

No Comments

No Known Activations