INDEX

Explanations

the repeated use of the token 'pa'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

SequentialGroup

-0.69

setViewportView

-0.57

 myſelf

-0.56

 itſelf

-0.51

 houſe

-0.50

 faſt

-0.49

 Shakspeare

-0.49

 poffe

-0.49

 ſta

-0.49

 ſte

-0.48

POSITIVE LOGITS

pa

0.66

 disambiguazione

0.65

 beast

0.57

 Beast

0.56

 slab

0.55

 hear

0.54

slab

0.52

Beast

0.51

Portail

0.49

 Generous

0.49

Activations Density 0.183%

the repeated use of the token 'pa'

No Comments

No Known Activations