INDEX

Explanations

studies and research findings discussing various topics

oai_token-act-pair · gpt-3.5-turbo

mentions of scientific studies or research findings

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

ward

-0.64

gr

-0.59

wards

-0.58

 blunt

-0.58

RAF

-0.57

eff

-0.57

polit

-0.57

nu

-0.57

IER

-0.57

 Postal

-0.55

POSITIVE LOGITS

uggest

1.08

 studies

1.02

study

1.00

udo

0.93

ilk

0.86

Study

0.83

heet

0.81

©¶æ

0.80

chool

0.78

ometimes

0.76

Activations Density 0.013%

studies and research findings discussing various topics

mentions of scientific studies or research findings

No Comments

No Known Activations