INDEX

Explanations

words related to knowledge, opinions, and beliefs

oai_token-act-pair · gpt-3.5-turbo

references to individuals or groups in statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

anmar

-0.72

 Recovery

-0.62

 Apocalypse

-0.56

roxy

-0.55

rug

-0.54

INAL

-0.53

exclusive

-0.53

annon

-0.53

>>>>>>>>

-0.53

 Mean

-0.52

POSITIVE LOGITS

 ever

1.15

soever

1.14

ever

1.09

pires

1.09

 imaginable

1.04

 EVER

1.04

 slightest

0.94

 touches

0.92

 interacts

0.89

 disagrees

0.88

Activations Density 0.112%

words related to knowledge, opinions, and beliefs

references to individuals or groups in statements

No Comments

No Known Activations