INDEX

Explanations

mentions of specific events or reports with emphasis, possibly headlines

oai_token-act-pair · gpt-3.5-turbo

titles or headings in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

cially

-0.73

 tremend

-0.72

alled

-0.71

Initialized

-0.69

uve

-0.66

entimes

-0.66

Ò

-0.66

ivable

-0.64

ounding

-0.64

rav

-0.64

POSITIVE LOGITS

How

0.96

 Latest

0.95

 https

0.94

 http

0.93

 Lessons

0.92

 Join

0.92

Who

0.90

Why

0.90

 Become

0.89

 Tips

0.88

Activations Density 0.134%

mentions of specific events or reports with emphasis, possibly headlines

titles or headings in a document

No Comments

No Known Activations