INDEX

Explanations

mentions of a location named "Hold" and variations of the word "old"

oai_token-act-pair · gpt-3.5-turbo

references to age or aging-related themes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 sclerosis

-0.82

 ILCS

-0.74

earchers

-0.73

kees

-0.72

ffee

-0.70

 mathemat

-0.70

ãĥ£

-0.69

pload

-0.68

aminer

-0.68

istine

-0.68

POSITIVE LOGITS

rums

1.07

est

0.97

sworth

0.92

 Trafford

0.90

ering

0.89

orf

0.89

face

0.87

stown

0.85

erer

0.84

enburg

0.84

Activations Density 0.038%

mentions of a location named "Hold" and variations of the word "old"

references to age or aging-related themes

No Comments

No Known Activations