INDEX

Explanations

phrases related to different types of professional activities or social issues

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a focus on performance or issues surrounding regulations

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

externalActionCode

-0.68

 thereof

-0.68

Cooldown

-0.63

Â·Â·

-0.59

Pic

-0.59

Ö

-0.58

Completed

-0.58

_.

-0.57

?".

-0.56

Chain

-0.56

POSITIVE LOGITS

 themselves

0.82

selves

0.76

iliar

0.67

 their

0.65

prisingly

0.63

raged

0.61

allery

0.59

enty

0.58

 puzzled

0.57

estern

0.56

Activations Density 1.208%

phrases related to different types of professional activities or social issues

phrases indicating a focus on performance or issues surrounding regulations

No Comments

No Known Activations