INDEX

Explanations

references to the word "Lie" or variations of it

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "lie."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

arthy

-0.77

oun

-0.73

smart

-0.72

iaries

-0.71

irlf

-0.70

atform

-0.67

uploads

-0.66

iles

-0.66

detail

-0.66

icable

-0.65

POSITIVE LOGITS

utenant

1.47

Lie

1.00

ge

0.94

uten

0.91

berman

0.88

ÃŁ

0.86

 detector

0.84

pard

0.83

Lie

0.83

yer

0.81

Activations Density 0.023%

references to the word "Lie" or variations of it

occurrences of the word "lie."

No Comments

No Known Activations