INDEX

Explanations

phrases related to performing tasks or actions

oai_token-act-pair · gpt-3.5-turbo

references to various tasks that can be performed

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

BN

-0.65

Esc

-0.62

 Mald

-0.60

 Civic

-0.60

STD

-0.57

ded

-0.57

LECT

-0.57

False

-0.57

ental

-0.56

Geo

-0.56

POSITIVE LOGITS

pring

1.29

afety

1.06

hops

1.02

 tasks

1.00

omething

0.93

uggest

0.92

mith

0.91

heet

0.91

paces

0.91

manship

0.87

Activations Density 0.031%

phrases related to performing tasks or actions

references to various tasks that can be performed

No Known Activations