INDEX

Explanations

mentions of specific sports and players

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Sports"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ipop

-0.76

idges

-0.74

uses

-0.71

iflower

-0.71

ered

-0.70

 miscar

-0.70

ĺħ

-0.68

utations

-0.68

 miscarriage

-0.67

yip

-0.66

POSITIVE LOGITS

 Sports

0.95

 Illustrated

0.93

 Entertainment

0.93

 Analytics

0.92

Center

0.85

Sports

0.84

Sport

0.83

 Club

0.82

Radio

0.81

sports

0.79

Activations Density 0.009%

mentions of specific sports and players

mentions of the term "Sports"

No Comments

No Known Activations