INDEX

Explanations

phrases related to additional or supplementary information in a given context

oai_token-act-pair · gpt-3.5-turbo

the word "also" to indicate additional information or points

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

jam

-0.69

atron

-0.61

behind

-0.60

 Commodore

-0.59

 Benson

-0.58

assic

-0.58

 Kinder

-0.58

 Breath

-0.57

Sm

-0.57

 Schedule

-0.57

POSITIVE LOGITS

been

1.03

 been

0.95

 undergone

0.86

cedented

0.81

 confir

0.80

 benefited

0.80

 volunte

0.78

yrinth

0.77

 fielded

0.77

ŃĶ

0.76

Activations Density 0.025%

phrases related to additional or supplementary information in a given context

the word "also" to indicate additional information or points

No Comments

No Known Activations