INDEX

Explanations

the word "Lie" or variations of it

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Lie" and variations of the name "Lieberman"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uploads

-0.71

ahime

-0.71

arthy

-0.71

runs

-0.67

anted

-0.67

 liking

-0.66

orsi

-0.65

ocked

-0.65

oulos

-0.65

âĶģ

-0.64

POSITIVE LOGITS

Lie

1.55

utenant

1.38

Lie

1.31

itle

0.93

lie

0.89

glers

0.84

gey

0.84

uten

0.82

ppel

0.82

zen

0.82

Activations Density 0.010%

the word "Lie" or variations of it

occurrences of the word "Lie" and variations of the name "Lieberman"

No Comments

No Known Activations