INDEX

Explanations

phrases containing the word "Ans"

oai_token-act-pair · gpt-3.5-turbo

references to characters or speakers named "Ans" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

iage

-0.70

 Manit

-0.68

dye

-0.66

 padd

-0.65

 deflation

-0.65

ttes

-0.64

 Haram

-0.64

 spac

-0.64

gging

-0.62

 captcha

-0.62

POSITIVE LOGITS

ans

0.84

weeney

0.76

lav

0.68

ullivan

0.67

Lee

0.66

anmar

0.65

essions

0.65

hen

0.65

Ang

0.63

he

0.63

Activations Density 0.003%

phrases containing the word "Ans"

references to characters or speakers named "Ans" in various contexts

No Comments

No Known Activations