INDEX

Explanations

scientific discoveries or breakthroughs

oai_token-act-pair · gpt-3.5-turbo

mentions of discoveries and findings

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

annis

-0.80

overe

-0.72

oiler

-0.69

osuke

-0.68

 paced

-0.67

cemic

-0.66

 bailed

-0.66

mph

-0.66

culated

-0.65

attery

-0.65

POSITIVE LOGITS

Ô

0.79

 Primordial

0.78

éļ

0.75

 Flavoring

0.73

iveness

0.73

Discover

0.72

 Wond

0.72

______

0.71

natureconservancy

0.71

âĸĴ

0.70

Activations Density 0.020%

scientific discoveries or breakthroughs

mentions of discoveries and findings

No Comments

No Known Activations