INDEX

Explanations

moral judgments and assessments of what is better or correct

oai_token-act-pair · gpt-3.5-turbo

ethical considerations and moral judgments in a discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Kings

-0.69

-0.61

 Ended

-0.58

CLASSIFIED

-0.56

asus

-0.54

ston

-0.53

 Ends

-0.52

phabet

-0.51

 Hits

-0.50

rongh

-0.50

POSITIVE LOGITS

to

0.99

for

0.88

 that

0.75

to

0.67

if

0.66

 practise

0.65

²¾

0.65

not

0.65

ceivable

0.65

 whether

0.62

Activations Density 0.150%

moral judgments and assessments of what is better or correct

ethical considerations and moral judgments in a discussion

No Comments

No Known Activations