INDEX

Explanations

references to countries or their details like registration, origin, or composition

oai_token-act-pair · gpt-3.5-turbo

references to countries and their associated metrics or characteristics

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

Charges

-0.73

mot

-0.69

atron

-0.67

Shift

-0.67

 Phantom

-0.67

Cop

-0.64

Mob

-0.64

Quotes

-0.63

tips

-0.61

crow

-0.61

POSITIVE LOGITS

 territories

0.85

 feder

0.82

 sovereign

0.79

 geographically

0.78

 sovereignty

0.78

lia

0.76

 populous

0.76

 federation

0.75

istan

0.75

States

0.74

Activations Density 0.564%

references to countries or their details like registration, origin, or composition

references to countries and their associated metrics or characteristics

No Comments

No Known Activations