INDEX

Explanations

proper nouns related to specific geographic locations

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names and locations related to a case or event

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Bav

-0.83

 Munich

-0.80

UV

-0.77

Um

-0.77

ept

-0.75

Naz

-0.71

 Caps

-0.71

ãģ®å®

-0.71

ulum

-0.70

 scen

-0.70

POSITIVE LOGITS

 Garfield

2.97

 Garner

2.31

Gar

1.81

 Ballard

1.60

 Randolph

1.51

 Monroe

1.25

cats

1.19

 Roku

1.02

 streetcar

1.01

 Barron

0.94

Activations Density 0.067%

proper nouns related to specific geographic locations

mentions of specific names and locations related to a case or event

No Comments

No Known Activations