INDEX

Explanations

instances where an action is noticed or observed

oai_token-act-pair · gpt-3.5-turbo

articles and demonstratives

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

 Jagu

-0.81

grounds

-0.80

Contents

-0.75

Edit

-0.70

 Amend

-0.68

 observes

-0.65

ie

-0.65

 assum

-0.65

align

-0.64

tests

-0.64

POSITIVE LOGITS

few

1.08

lot

1.06

 plethora

1.05

 handful

1.04

 glimpse

1.03

 bunch

1.01

 significant

0.98

 multitude

0.98

 huge

0.98

 couple

0.96

Activations Density 0.433%

instances where an action is noticed or observed

articles and demonstratives

No Comments

No Known Activations