INDEX

Explanations

references to Osama bin Laden

oai_token-act-pair · gpt-3.5-turbo

references to Osama bin Laden

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

Medic

-0.68

eries

-0.68

rep

-0.67

 Constantin

-0.67

chrom

-0.66

ISS

-0.65

perty

-0.63

Wide

-0.63

fx

-0.62

 DRAG

-0.62

POSITIVE LOGITS

 Laden

1.52

abad

0.91

bin

0.87

ghan

0.87

gebra

0.87

ãĥĥãĥī

0.87

querque

0.85

addafi

0.85

 Jinping

0.81

iyah

0.79

Activations Density 0.010%

references to Osama bin Laden

references to Osama bin Laden

No Comments

No Known Activations