INDEX

Explanations

phrases related to investigating, exploring, or delving into various subjects or situations

oai_token-act-pair · gpt-3.5-turbo

themes of digging and lifting 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

aughtered

-0.76

exist

-0.68

aceae

-0.67

iste

-0.66

vez

-0.65

 Changed

-0.63

izons

-0.63

ursed

-0.62

=-=-=-=-=-=-=-=-

-0.62

iren

-0.61

POSITIVE LOGITS

 tearing

2.05

 splitting

1.95

 digging

1.92

 swapping

1.91

 ripping

1.89

 tapping

1.87

 squeezing

1.86

 stripping

1.85

 shedding

1.85

 lifting

1.84

Activations Density 0.115%

phrases related to investigating, exploring, or delving into various subjects or situations

themes of digging and lifting in various contexts

No Comments

No Known Activations