INDEX

Explanations

words related to justice or justifications

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "just" and its variations, indicating a focus on justifications or fairness

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

 cous

-0.72

luster

-0.69

uberty

-0.69

yip

-0.68

xual

-0.67

pora

-0.66

 Prelude

-0.66

wcs

-0.63

 monop

-0.61

yang

-0.61

POSITIVE LOGITS

ifiable

1.40

ifications

1.33

ices

1.12

if

1.01

ifi

0.98

ified

0.97

icia

0.93

ifiers

0.93

ification

0.93

IFIED

0.92

Activations Density 0.087%

words related to justice or justifications

occurrences of the word "just" and its variations, indicating a focus on justifications or fairness

No Comments

No Known Activations