INDEX

Explanations

comparisons between different things

oai_token-act-pair · gpt-3.5-turbo

comparisons that emphasize something's superiority in relation to another

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

estern

-0.74

illary

-0.73

osen

-0.68

auri

-0.67

onomy

-0.67

umbn

-0.67

pez

-0.66

auer

-0.66

ãĤ¨

-0.65

kan

-0.65

POSITIVE LOGITS

 anything

1.02

 necessarily

0.96

any

0.74

 outright

0.70

nam

0.69

 condone

0.66

 adequately

0.66

 substance

0.66

 substantive

0.65

 merely

0.64

Activations Density 0.105%

comparisons between different things

comparisons that emphasize something's superiority in relation to another

No Comments

No Known Activations