INDEX

Explanations

terms and phrases related to the classification and definitions of groups or labels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

featureID

-0.56

majánló

-0.56

Vidite

-0.55

 ModelRenderer

-0.54

SequentialGroup

-0.52

InitVars

-0.51

+#+#

-0.50

StructEnd

-0.50

ſammen

-0.49

 Fines

-0.48

POSITIVE LOGITS

 term

0.45

Tikang

0.44

 Begriffe

0.41

 terms

0.41

 istilah

0.39

 phrases

0.39

 acronyms

0.39

 terminology

0.39

 phrase

0.38

term

0.36

Activations Density 0.461%

terms and phrases related to the classification and definitions of groups or labels

No Comments

No Known Activations