INDEX

Explanations

proper nouns related to people or companies

oai_token-act-pair · gpt-3.5-turbo

terms related to entities or characters, particularly those containing the prefix "Ax."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ished

-0.70

asted

-0.66

ufact

-0.66

merce

-0.66

lights

-0.65

committee

-0.65

atural

-0.65

lighting

-0.64

 checkout

-0.62

soDeliveryDate

-0.60

POSITIVE LOGITS

iom

1.04

xon

1.02

ios

0.93

xes

0.92

illary

0.88

AX

0.87

ansas

0.86

Ax

0.85

Ax

0.84

iop

0.83

Activations Density 0.006%

proper nouns related to people or companies

terms related to entities or characters, particularly those containing the prefix "Ax."

No Comments

No Known Activations