INDEX

Explanations

numbers followed by a mix of letters and symbols

oai_token-act-pair · gpt-3.5-turbo

many numerical values associated with various technical specifications or models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 snowball

-0.74

querque

-0.68

 volunt

-0.66

selves

-0.65

 polic

-0.64

 Mexicans

-0.64

 lawy

-0.64

 ourselves

-0.64

âĸ¬

-0.63

 yourselves

-0.61

POSITIVE LOGITS

 chipset

1.03

RG

1.00

TX

1.00

UF

1.00

MK

1.00

PF

1.00

LR

1.00

CV

0.97

DM

0.97

DX

0.97

Activations Density 0.096%

numbers followed by a mix of letters and symbols

many numerical values associated with various technical specifications or models

No Comments

No Known Activations