INDEX

Explanations

proper nouns and titles of various entities

oai_token-act-pair · gpt-3.5-turbo

articles that signal the introduction of nouns or noun phrases

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

anism

-0.78

 Izan

-0.76

 chops

-0.75

 denies

-0.75

 knows

-0.73

 acknowledges

-0.71

chuk

-0.70

faces

-0.68

could

-0.68

 listens

-0.68

POSITIVE LOGITS

 versatile

0.89

 unique

0.83

 fascinating

0.83

 prominent

0.83

 standalone

0.82

 significant

0.82

 comprehensive

0.81

igmatic

0.80

 wonderful

0.80

 huge

0.80

Activations Density 0.187%

proper nouns and titles of various entities

articles that signal the introduction of nouns or noun phrases

No Comments

No Known Activations