INDEX

Explanations

phrases with the word "big"

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "big" in various contexts

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

 briefly

-0.70

 temporarily

-0.69

 respectively

-0.66

 rendered

-0.66

 restored

-0.65

med

-0.62

 para

-0.62

 momentarily

-0.62

 subsequently

-0.62

 notwithstanding

-0.61

POSITIVE LOGITS

big

3.89

Big

2.25

largest

1.78

huge

1.77

small

1.73

BIG

1.69

big

1.56

Big

1.49

 bigger

1.42

little

1.36

Activations Density 0.009%

phrases with the word "big"

repeated mentions of the word "big" in various contexts

No Comments

No Known Activations