INDEX

Explanations

references to winning or receiving trophies

oai_token-act-pair · gpt-3.5-turbo

mentions of trophies

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

utenberg

-0.80

 hypers

-0.71

 mixing

-0.68

ket

-0.66

 Zeit

-0.66

trak

-0.64

 propensity

-0.64

 sketch

-0.63

andem

-0.62

weeney

-0.61

POSITIVE LOGITS

 trophy

1.45

 trophies

1.40

 Trophy

1.34

phies

1.13

abilia

0.74

ãĤ©

0.74

brates

0.74

ilage

0.74

 vanquished

0.73

BUG

0.73

Activations Density 0.004%

references to winning or receiving trophies

mentions of trophies

No Comments

No Known Activations