INDEX

Explanations

proper nouns and terms that are being credited for something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "credited" and its variations

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

xy

-0.71

FO

-0.68

 tents

-0.67

cre

-0.61

por

-0.61

xon

-0.60

perty

-0.60

adra

-0.59

psy

-0.59

 Afgh

-0.59

POSITIVE LOGITS

 credited

1.12

rill

0.98

olor

0.95

 credits

0.91

redited

0.86

 Credits

0.79

rious

0.77

heet

0.76

ouble

0.76

Credit

0.75

Activations Density 0.010%

proper nouns and terms that are being credited for something

instances of the word "credited" and its variations

No Comments

No Known Activations