INDEX

Explanations

references to sports-related content

oai_token-act-pair · gpt-3.5-turbo

references to sports-related content

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

izons

-0.74

 presumptive

-0.66

 unborn

-0.66

umps

-0.64

 convincing

-0.64

ricanes

-0.64

 decipher

-0.63

 ç¥ŀ

-0.62

 sidx

-0.61

haul

-0.61

POSITIVE LOGITS

Sport

1.35

 Sport

1.24

 Wrestling

0.94

sports

0.93

scar

0.93

club

0.91

sw

0.86

 Club

0.85

bike

0.84

Club

0.83

Activations Density 0.006%

references to sports-related content

references to sports-related content

No Comments

No Known Activations