INDEX

Explanations

mentions of geographical locations, specifically focusing on the city of Tehran

oai_token-act-pair · gpt-3.5-turbo

mentions of specific geographical locations, particularly Tehran and Ankara

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Uncommon

-0.84

ICAN

-0.83

oslov

-0.79

ONSORED

-0.74

 Crus

-0.72

ESH

-0.70

ICO

-0.69

ATHER

-0.69

VPN

-0.68

estern

-0.68

POSITIVE LOGITS

ascus

1.20

hran

0.97

 Lumpur

0.97

 Ankara

0.89

 rall

0.87

 Islamabad

0.85

mouth

0.83

pole

0.78

ffiti

0.76

ashtra

0.75

Activations Density 0.013%

mentions of geographical locations, specifically focusing on the city of Tehran

mentions of specific geographical locations, particularly Tehran and Ankara

No Comments

No Known Activations