INDEX

Explanations

references to specific skills or abilities

oai_token-act-pair · gpt-3.5-turbo

references to levels of skill and expertise

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

Jane

-0.71

 PubMed

-0.68

 Amen

-0.65

Wem

-0.64

 Corinthians

-0.64

 Slate

-0.64

 Persian

-0.64

 Sawyer

-0.63

DSM

-0.63

 Pall

-0.62

POSITIVE LOGITS

anooga

1.01

 skill

0.95

smanship

0.93

Skill

0.93

poke

0.92

 skills

0.91

fully

0.88

ful

0.86

uracy

0.81

 veter

0.79

Activations Density 0.007%

references to specific skills or abilities

references to levels of skill and expertise

No Comments

No Known Activations