INDEX

Explanations

phrases related to blog posts or written articles

oai_token-act-pair · gpt-3.5-turbo

references to blog posts

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

estial

-0.92

ãĤ¤ãĥĪ

-0.78

ichick

-0.76

 Orchestra

-0.75

isin

-0.67

 Flavoring

-0.67

 Beans

-0.67

 Dise

-0.65

owship

-0.65

 bands

-0.64

POSITIVE LOGITS

gres

1.26

doctoral

1.15

mortem

1.09

doc

0.98

hum

0.91

natal

0.91

docs

0.88

mort

0.88

hoc

0.87

greSQL

0.86

Activations Density 0.037%

phrases related to blog posts or written articles

references to blog posts

No Comments

No Known Activations