INDEX

Explanations

keywords associated with the name "JP"

oai_token-act-pair · gpt-3.5-turbo

references to financial institutions and corporate entities, particularly JPMorgan

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 thirds

-0.80

rums

-0.79

querque

-0.77

rum

-0.73

 Rebels

-0.70

rights

-0.68

taboola

-0.68

 Masquerade

-0.65

byss

-0.64

leneck

-0.64

POSITIVE LOGITS

oint

1.03

JP

1.00

enny

0.96

avascript

0.95

uty

0.93

acket

0.89

ealous

0.88

ython

0.87

oha

0.87

ACK

0.87

Activations Density 0.018%

keywords associated with the name "JP"

references to financial institutions and corporate entities, particularly JPMorgan

No Comments

No Known Activations