INDEX

Explanations

phrases related to evaluation or assessment

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the existence or importance of certain ideas or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ´ãĥ³

-0.84

iren

-0.82

aired

-0.81

urred

-0.76

aughtered

-0.75

retched

-0.74

izens

-0.74

pered

-0.73

affiliated

-0.72

rup

-0.69

POSITIVE LOGITS

 gonna

0.98

 figuring

0.97

 trying

0.86

 getting

0.85

how

0.81

 consistency

0.81

 educating

0.80

 putting

0.80

 thanking

0.79

 respecting

0.79

Activations Density 0.265%

phrases related to evaluation or assessment

phrases indicating the existence or importance of certain ideas or actions

No Comments

No Known Activations