INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

names of key individuals mentioned in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

bar

-0.69

ger

-0.69

hot

-0.65

space

-0.64

jab

-0.63

inic

-0.62

Applications

-0.62

JP

-0.62

 recl

-0.62

abad

-0.61

POSITIVE LOGITS

 Rouse

2.16

 Simpson

1.92

 Dodd

1.63

 Dunn

1.59

 Nunes

1.22

 Walton

1.11

FI

1.06

Lum

1.04

 Commissioners

1.02

 Campbell

0.98

Activations Density 0.022%

names of individuals

names of key individuals mentioned in the text

No Comments

No Known Activations