INDEX

Explanations

words containing the substring "irm"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "affirm" and its derivatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

REDACTED

-0.75

ç«

-0.67

å¹

-0.66

 Cong

-0.63

Rus

-0.62

å¥

-0.62

 largeDownload

-0.62

Ree

-0.62

lihood

-0.62

HY

-0.61

POSITIVE LOGITS

irm

1.75

irms

1.10

irmed

1.00

ament

0.99

atively

0.97

etric

0.95

irmation

0.91

aton

0.91

ative

0.89

ities

0.86

Activations Density 0.007%

words containing the substring "irm"

variations of the word "affirm" and its derivatives

No Comments

No Known Activations