INDEX

Explanations

Japanese names from different fields like entertainment and technology

oai_token-act-pair · gpt-3.5-turbo

names and terms related to gaming and entertainment

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

 coinc

-0.71

CFR

-0.70

 crossings

-0.66

lus

-0.65

 advertis

-0.64

bc

-0.64

 android

-0.63

AFL

-0.63

 markets

-0.62

 comp

-0.62

POSITIVE LOGITS

iken

3.60

oki

3.38

ickson

1.29

oka

1.12

iky

1.08

ahi

1.04

uum

1.00

efer

0.96

odon

0.89

oha

0.88

Activations Density 0.031%

Japanese names from different fields like entertainment and technology

names and terms related to gaming and entertainment

No Comments

No Known Activations