INDEX

Explanations

words related to competition and battles for positions or titles

oai_token-act-pair · gpt-3.5-turbo

phrases related to competitive events or contests

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

raq

-0.87

ires

-0.80

reet

-0.70

ighed

-0.69

quist

-0.69

reed

-0.68

cano

-0.68

ower

-0.68

fter

-0.68

Normal

-0.65

POSITIVE LOGITS

 supremacy

1.37

 dominance

1.09

 bragging

1.08

 throne

1.03

 spots

1.02

 titles

0.99

 superiority

0.98

 championships

0.97

 fame

0.96

 hearts

0.92

Activations Density 0.170%

words related to competition and battles for positions or titles

phrases related to competitive events or contests

No Comments

No Known Activations