INDEX

Explanations

any references to physical objects

oai_token-act-pair · gpt-3.5-turbo

references to various objects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

tes

-0.70

BILITY

-0.67

neck

-0.63

 nerve

-0.63

 Hampton

-0.63

ther

-0.63

teenth

-0.62

ceed

-0.62

OUNT

-0.61

 millenn

-0.60

POSITIVE LOGITS

objects

1.09

 objects

1.08

wana

0.90

uggest

0.88

matter

0.86

ĸļ

0.85

*/(

0.84

imus

0.82

 Objects

0.80

istence

0.80

Activations Density 0.010%

any references to physical objects

references to various objects

No Comments

No Known Activations