INDEX

Explanations

technical terms related to specific capabilities or attributes

oai_token-act-pair · gpt-3.5-turbo

phrases related to attributes and characteristics of various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Worse

-0.84

emies

-0.81

ogs

-0.79

UGH

-0.77

gey

-0.75

agues

-0.74

 Horowitz

-0.73

 Journals

-0.73

hemy

-0.71

acers

-0.71

POSITIVE LOGITS

 versatility

1.51

 simplicity

1.47

 flexibility

1.30

 unparalleled

1.28

 ease

1.26

 uniqueness

1.26

 unmatched

1.25

 proximity

1.24

 affordability

1.24

 ability

1.21

Activations Density 0.336%

technical terms related to specific capabilities or attributes

phrases related to attributes and characteristics of various subjects

No Comments

No Known Activations