INDEX

Explanations

mentions of media organizations and news-related terms

oai_token-act-pair · gpt-3.5-turbo

words and abbreviations relevant to media organizations and journalistic sources

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

)",

-0.61

 Valve

-0.58

Magikarp

-0.55

 franchise

-0.54

 setting

-0.53

 fame

-0.53

 decomp

-0.53

 dressing

-0.53

 Borderlands

-0.52

 colleg

-0.51

POSITIVE LOGITS

 WATCHED

1.01

 guiActiveUn

0.87

 UNCLASSIFIED

0.77

NEWS

0.75

 PHOTO

0.70

<|endoftext|>

0.70

UFF

0.70

 Shutterstock

0.69

 Photograph

0.68

 SHARES

0.67

Activations Density 0.515%

mentions of media organizations and news-related terms

words and abbreviations relevant to media organizations and journalistic sources

No Comments

No Known Activations