INDEX

Explanations

news agencies or publishing entities

oai_token-act-pair · gpt-3.5-turbo

mentions of news agencies or media outlets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

profits

-0.67

abiding

-0.64

 forgiven

-0.62

âĹ¼

-0.61

 advis

-0.60

table

-0.59

conservancy

-0.58

righteous

-0.56

 concluding

-0.56

kept

-0.56

POSITIVE LOGITS

 IMAGES

1.09

 Images

0.99

Images

0.97

 PHOTO

0.96

File

0.90

 FILE

0.89

photo

0.83

 Photos

0.83

0.79

reenshot

0.78

Activations Density 0.040%

news agencies or publishing entities

mentions of news agencies or media outlets

No Comments

No Known Activations