INDEX

Explanations

keywords related to custom products or services

oai_token-act-pair · gpt-3.5-turbo

references to customization and bewilderment

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

Not in Any Lists

No Comments

Negative Logits

hered

-0.79

agnar

-0.66

 soluble

-0.64

 NOTICE

-0.64

SI

-0.62

clinton

-0.60

channelAvailability

-0.60

 Dame

-0.60

ED

-0.59

dos

-0.59

POSITIVE LOGITS

ers

1.99

ership

1.90

er

1.73

ered

1.12

ery

1.11

ering

1.03

ER

1.02

ern

1.00

erers

0.97

ERS

0.97

Activations Density 0.041%

keywords related to custom products or services

references to customization and bewilderment

No Comments

No Known Activations