INDEX

Explanations

phrases related to news and alerts

oai_token-act-pair · gpt-3.5-turbo

references to breaking news or urgent updates

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

haps

-0.82

inating

-0.78

ateurs

-0.76

inators

-0.74

ateur

-0.74

igmat

-0.71

itism

-0.71

Ã¨re

-0.70

uther

-0.69

yon

-0.69

POSITIVE LOGITS

fast

0.92

 news

0.85

 NEWS

0.79

AKING

0.78

 revelations

0.71

neck

0.68

 News

0.67

 EVENTS

0.66

 loose

0.66

Berry

0.65

Activations Density 0.025%

phrases related to news and alerts

references to breaking news or urgent updates

No Comments

No Known Activations