INDEX

Explanations

terms related to document layout or design

oai_token-act-pair · gpt-3.5-turbo

references to various types of layouts

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

icer

-0.99

jamin

-0.92

hood

-0.91

inas

-0.77

elope

-0.74

 Lauder

-0.74

iveness

-0.74

lehem

-0.73

Ago

-0.71

nard

-0.70

POSITIVE LOGITS

 layouts

1.28

 layout

1.19

 Layout

1.07

layout

0.88

 diagrams

0.84

 mismatch

0.82

 diagram

0.80

Layout

0.79

 spacing

0.79

xml

0.76

Activations Density 0.012%

terms related to document layout or design

references to various types of layouts

No Comments

No Known Activations