INDEX

Explanations

words representing emphasis or focus

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "word" and its relevance in 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Farms

-0.67

leton

-0.66

 Supports

-0.66

cade

-0.66

strate

-0.65

alia

-0.64

elson

-0.64

eers

-0.61

erton

-0.61

 Paddock

-0.60

POSITIVE LOGITS

 word

3.87

word

2.51

Word

2.34

 Word

2.19

 words

2.03

 phrase

1.83

words

1.65

 term

1.62

 adjective

1.54

 Words

1.50

Activations Density 0.014%

words representing emphasis or focus

occurrences of the word "word" and its relevance in context

No Comments

No Known Activations