INDEX

Explanations

references to data and information in a structured or academic context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 Réponses

-0.83

 calendriers

-0.76

NewUrlParser

-0.65

ongiorno

-0.65

 tartalomajánló

-0.63

min

-0.62

تقاوى

-0.59

hitheatre

-0.59

-0.58

tymologie

-0.56

POSITIVE LOGITS

withstanding

0.99

Efq

0.97

 Perſ

0.90

 Theſe

0.84

ſelf

0.83

 Majefty

0.83

 pleaſure

0.82

 Monfieur

0.82

 JSTOR

0.80

 Conſ

0.78

Activations Density 1.600%

references to data and information in a structured or academic context

No Comments

No Known Activations