INDEX

Explanations

references to the Super Smash Bros video game series

oai_token-act-pair · gpt-3.5-turbo

references to the Super Smash Bros. video game franchise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 foss

-0.74

judicial

-0.66

rises

-0.66

 Lumpur

-0.63

 Minotaur

-0.63

otype

-0.62

 sequest

-0.61

urion

-0.61

 separ

-0.61

 pigment

-0.60

POSITIVE LOGITS

halla

1.08

words

0.87

 Bros

0.87

bowl

0.82

 Smash

0.78

adelphia

0.76

 Mouth

0.74

 Brawl

0.73

 Wrestling

0.71

jam

0.70

Activations Density 0.020%

references to the Super Smash Bros video game series

references to the Super Smash Bros. video game franchise

No Comments

No Known Activations