INDEX

Explanations

words related to claims or assertions about identity or classification

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

andaag

-0.50

ագրություններ

-0.49

новништво

-0.48

knapp

-0.48

новниш

-0.48

 mapStateToProps

-0.48

 handicap

-0.46

ăn

-0.44

 beat

-0.43

 Vikipedi

-0.42

POSITIVE LOGITS

 possibile

0.73

 presumed

0.63

 putative

0.62

ORMAL

0.62

 called

0.62

 quelcon

0.61

 Called

0.61

 Monfieur

0.60

 Chriſt

0.60

__*/

0.60

Activations Density 0.470%

words related to claims or assertions about identity or classification

No Comments

No Known Activations