INDEX

Explanations

phrases related to understanding and interpretation

oai_token-act-pair · gpt-3.5-turbo

references to understanding complex concepts or questions

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

icides

-0.84

icide

-0.75

exclusive

-0.74

pload

-0.71

olphins

-0.70

veyard

-0.70

odder

-0.69

uable

-0.69

uxe

-0.67

icho

-0.66

POSITIVE LOGITS

 workings

0.81

 Centauri

0.73

 nuances

0.73

 plight

0.71

 limitations

0.71

ĨĴ

0.69

 concepts

0.68

 intric

0.67

 lesson

0.65

 Keynes

0.65

Activations Density 0.192%

phrases related to understanding and interpretation

references to understanding complex concepts or questions

No Comments

No Known Activations