INDEX

Explanations

phrases or words that attribute something to a specific cause or source

oai_token-act-pair · gpt-3.5-turbo

the word "attributed" and its variations 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

corn

-0.82

estern

-0.75

aspx

-0.73

 Sandwich

-0.73

lake

-0.70

 Stras

-0.69

atoes

-0.69

 Maple

-0.68

xtap

-0.67

stall

-0.66

POSITIVE LOGITS

 attribut

1.02

 blame

0.98

 attribution

0.85

 attributed

0.80

 forgiven

0.79

 attributable

0.76

 corrid

0.76

imir

0.75

 unden

0.75

illac

0.72

Activations Density 0.012%

phrases or words that attribute something to a specific cause or source

the word "attributed" and its variations in various contexts

No Comments

No Known Activations