INDEX

Explanations

words related to awards, competitions, and achievements

oai_token-act-pair · gpt-3.5-turbo

references to trophies and awards

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

weeney

-0.77

duino

-0.73

utenberg

-0.72

 experimenting

-0.72

pace

-0.68

lement

-0.66

trak

-0.65

erman

-0.64

oscope

-0.64

Osw

-0.64

POSITIVE LOGITS

 Trophy

1.11

 trophies

1.05

 trophy

1.04

phies

0.97

brates

0.82

 Rica

0.80

brate

0.76

ilage

0.73

aceous

0.72

cake

0.71

Activations Density 0.021%

words related to awards, competitions, and achievements

references to trophies and awards

No Comments

No Known Activations