INDEX

Explanations

phrases related to targets or objectives

oai_token-act-pair · gpt-3.5-turbo

phrases related to intention or purpose

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

--------------------------------------------------------

-0.76

PASS

-0.76

Redd

-0.72

Column

-0.71

------------------------------------------------

-0.70

 Cust

-0.69

 Solitaire

-0.69

NetMessage

-0.65

Berry

-0.64

 Roses

-0.64

POSITIVE LOGITS

aim

1.06

yip

0.99

lessly

0.94

gha

0.88

atory

0.86

emouth

0.85

rious

0.84

folios

0.83

lessness

0.82

 aims

0.82

Activations Density 0.005%

phrases related to targets or objectives

phrases related to intention or purpose

No Comments

No Known Activations