INDEX

Explanations

words indicating a change or transition from one state or condition to another

oai_token-act-pair · gpt-3.5-turbo

transitions represented by the word 'from'

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

iosyn

-0.84

vic

-0.82

QUI

-0.80

mask

-0.76

successfully

-0.76

iquette

-0.75

iar

-0.71

VIDEO

-0.71

leneck

-0.71

imentary

-0.70

POSITIVE LOGITS

 afar

1.11

 thence

0.85

 abroad

0.83

 whence

0.83

 indoors

0.75

 obscurity

0.74

 somewhere

0.74

 upstairs

0.71

 across

0.71

 downstairs

0.69

Activations Density 0.096%

words indicating a change or transition from one state or condition to another

transitions represented by the word 'from'

No Comments

No Known Activations