INDEX

Explanations

phrases related to leadership positions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 revolves

-0.66

repeat

-0.65

 because

-0.65

 redund

-0.63

 masturb

-0.63

 favour

-0.63

 collide

-0.62

coins

-0.62

 whilst

-0.61

NULL

-0.61

POSITIVE LOGITS

 aforementioned

0.98

 National

0.92

 Institute

0.89

 latter

0.86

 Department

0.85

 United

0.83

 International

0.81

 Office

0.81

 same

0.80

 largest

0.80

Activations Density 0.370%

phrases related to leadership positions

instances of the word "the."

No Comments

No Known Activations