INDEX

Explanations

abbreviations or mentions of "BC" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to British Columbia (BC) in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

icio

-0.80

hyde

-0.71

zie

-0.67

dir

-0.65

sein

-0.64

 embargo

-0.64

 marsh

-0.63

ennis

-0.62

stri

-0.62

 arsen

-0.62

POSITIVE LOGITS

BC

3.84

BC

2.21

BCE

2.20

bc

1.78

CE

1.49

BCC

1.46

NSW

1.44

bc

1.42

BD

1.34

 Manitoba

1.34

Activations Density 0.014%

abbreviations or mentions of "BC" followed by a number

references to British Columbia (BC) in various contexts

No Comments

No Known Activations