INDEX

Explanations

the name "ke" in various contexts

oai_token-act-pair · gpt-3.5-turbo

the word "ke" and its variations in different contexts

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

 Palestin

-0.76

displayText

-0.71

NRS

-0.68

ourcing

-0.67

ourced

-0.66

 apprehension

-0.64

ISO

-0.64

tested

-0.63

 scrimmage

-0.63

ologically

-0.63

POSITIVE LOGITS

nda

1.17

llers

1.04

ke

1.04

ptic

1.03

ller

1.01

ffe

0.98

leys

0.98

jriwal

0.97

les

0.92

vich

0.92

Activations Density 0.008%

the name "ke" in various contexts

the word "ke" and its variations in different contexts

No Comments

No Known Activations