INDEX

Explanations

references to the sports team "Spurs"

oai_token-act-pair · gpt-3.5-turbo

mentions of the Spurs basketball team

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ROM

-0.71

 cycl

-0.66

MM

-0.65

oros

-0.64

NA

-0.63

mah

-0.63

gl

-0.63

ms

-0.63

vernment

-0.62

BJ

-0.62

POSITIVE LOGITS

 Spurs

3.61

 Tottenham

2.66

ottenham

2.21

Poc

1.61

 Spur

1.54

 Hots

1.47

 Pistons

1.43

 Chargers

1.41

 Grizzlies

1.40

 Arsenal

1.36

Activations Density 0.014%

references to the sports team "Spurs"

mentions of the Spurs basketball team

No Comments

No Known Activations