INDEX

Explanations

instances of the word "nam" in different contexts

oai_token-act-pair · gpt-3.5-turbo

occurrences of names and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Introduced

-0.67

 Blackwell

-0.63

haz

-0.62

ECB

-0.60

 respect

-0.60

 understatement

-0.59

 forecast

-0.58

UID

-0.58

 cider

-0.57

wings

-0.56

POSITIVE LOGITS

nam

1.34

orously

1.08

borgh

1.03

ned

1.03

eless

1.02

aji

0.96

aceutical

0.95

oji

0.95

entary

0.95

pton

0.94

Activations Density 0.005%

instances of the word "nam" in different contexts

occurrences of names and related terms

No Comments

No Known Activations