INDEX

Explanations

mentions of media, news, and public attention

oai_token-act-pair · gpt-3.5-turbo

references to various forms of media and public reactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

HI

-0.64

icidal

-0.61

CHA

-0.61

TPS

-0.61

Instruct

-0.59

 subsistence

-0.59

 administr

-0.59

Viol

-0.58

 Shutterstock

-0.58

Dialog

-0.58

POSITIVE LOGITS

 clam

1.06

 pundits

1.03

 understandably

1.02

 alike

0.98

 outcry

0.94

 eagerly

0.94

 acclaim

0.93

 speculate

0.93

 speculated

0.92

 scorn

0.91

Activations Density 0.458%

mentions of media, news, and public attention

references to various forms of media and public reactions

No Comments

No Known Activations