INDEX

Explanations

editorial notes or annotations within a text

oai_token-act-pair · gpt-3.5-turbo

references to editorial notes or comments from editors

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

ãĥĺ

-0.73

unin

-0.71

ãģ®éŃĶ

-0.69

asar

-0.67

milo

-0.67

arians

-0.66

naires

-0.65

 vain

-0.65

oner

-0.63

vas

-0.63

POSITIVE LOGITS

 Note

1.41

 Picks

1.35

 note

1.26

 Rating

1.14

 Notes

1.14

 Comment

1.13

 Recommend

1.10

 Comments

1.08

 NOTE

1.04

Note

1.01

Activations Density 0.087%

editorial notes or annotations within a text

references to editorial notes or comments from editors

No Comments

No Known Activations