INDEX

Explanations

links to additional information or instructions

oai_token-act-pair · gpt-3.5-turbo

references to articles and tutorials for further information

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

yss

-0.92

phony

-0.72

stocks

-0.71

avorite

-0.70

âĻ¦

-0.68

orescence

-0.67

 unthinkable

-0.67

Ú

-0.67

icides

-0.67

Īè

-0.66

POSITIVE LOGITS

 documentation

1.73

 docs

1.69

 Documentation

1.50

 wiki

1.46

 tutorials

1.40

 tutorial

1.39

FAQ

1.35

 guides

1.35

 guide

1.32

 manual

1.30

Activations Density 0.344%

links to additional information or instructions

references to articles and tutorials for further information

No Comments

No Known Activations