INDEX

Explanations

specific acronyms and technical terms

oai_token-act-pair · gpt-3.5-turbo

parenthetical expressions and acronyms, indicating the presence of structured or technical information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

KH

-0.89

 puff

-0.77

HG

-0.77

DH

-0.77

WB

-0.77

BW

-0.76

HB

-0.75

AQ

-0.75

SB

-0.74

DU

-0.73

POSITIVE LOGITS

International

1.38

formerly

1.38

Advanced

1.33

pron

1.33

Secure

1.30

Universal

1.29

Independent

1.25

Elect

1.24

National

1.24

Personal

1.24

Activations Density 0.071%

specific acronyms and technical terms

parenthetical expressions and acronyms, indicating the presence of structured or technical information

No Comments

No Known Activations