INDEX

Explanations

mention or refer to Lebanon or entities related to Lebanon

oai_token-act-pair · gpt-3.5-turbo

references to Lebanon and its political or military context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

gers

-0.92

 Ultron

-0.78

VERTISEMENT

-0.72

mson

-0.69

earable

-0.66

 sequ

-0.65

zag

-0.64

 geek

-0.64

 Clicker

-0.64

 Ducks

-0.63

POSITIVE LOGITS

 Hezbollah

1.42

 Beirut

1.30

 Lebanon

1.26

 Lebanese

1.22

 Leban

1.21

 Shiite

1.06

raq

0.94

oun

0.85

 Kurd

0.82

Mub

0.82

Activations Density 0.039%

mention or refer to Lebanon or entities related to Lebanon

references to Lebanon and its political or military context

No Comments

No Known Activations