INDEX

Explanations

specific types of structured text or code elements, along with comparisons in scientific contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

böz

-0.51

 tủ

-0.48

 secco

-0.48

 Infórmanos

-0.47

 ModelExpression

-0.46

 dormitórios

-0.44

roep

-0.43

 pittores

-0.43

ynchronously

-0.43

在下

-0.43

POSITIVE LOGITS

errHandler

0.66

 Monfieur

0.57

 Viel

0.55

::::::::

0.53

 édition

0.53

mphony

0.53

ρας

0.52

Hochspringen

0.52

 perle

0.51

({},

0.50

Activations Density 0.057%

specific types of structured text or code elements, along with comparisons in scientific contexts

No Comments

No Known Activations