INDEX

Explanations

the word "alo" with different activations, especially "alo" with high activation

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "alo" or related variations in context

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

dayName

-0.66

 Bedford

-0.66

 Consent

-0.64

 Debor

-0.63

Present

-0.63

stood

-0.63

ienced

-0.63

 Schumer

-0.62

 Berks

-0.62

 Clause

-0.62

POSITIVE LOGITS

alo

1.26

axy

1.08

osa

0.97

zona

0.95

ose

0.92

axies

0.91

osity

0.88

iety

0.86

orthy

0.85

oso

0.85

Activations Density 0.003%

the word "alo" with different activations, especially "alo" with high activation

occurrences of the term "alo" or related variations in context

No Comments

No Known Activations