INDEX

Explanations

the word "a"

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing justice and legal concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

wake

-0.62

atron

-0.60

 Jarrett

-0.58

 Purg

-0.56

Von

-0.56

ãĢĲ

-0.55

 Randy

-0.54

 Vest

-0.53

urities

-0.53

 Stras

-0.53

POSITIVE LOGITS

 sqor

0.75

 tremend

0.66

umatic

0.61

 Means

0.59

xual

0.58

istant

0.58

 resemblance

0.58

itely

0.58

 abbre

0.57

utenberg

0.56

Activations Density 1.088%

the word "a"

phrases emphasizing justice and legal concepts

No Comments

No Known Activations