INDEX

Explanations

phrases beginning with 'Each'

oai_token-act-pair · gpt-3.5-turbo

references to individual contributions within a collective context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 quite

-0.68

 anyway

-0.65

 embarrass

-0.61

 negative

-0.60

 freak

-0.59

 offensive

-0.58

 misinterpret

-0.58

 sentiment

-0.58

 very

-0.58

 probably

-0.58

POSITIVE LOGITS

 Each

3.22

Each

2.36

each

2.27

 each

1.90

 Every

1.85

 Whenever

1.44

Every

1.42

 Both

1.40

 apiece

1.36

 These

1.32

Activations Density 0.014%

phrases beginning with 'Each'

references to individual contributions within a collective context

No Comments

No Known Activations