INDEX

Explanations

adjectives followed by "of" and then a noun

oai_token-act-pair · gpt-3.5-turbo

phrases that categorize or describe types of entities or concepts

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

mercial

-0.78

 VIDEOS

-0.78

abases

-0.76

plings

-0.75

ULTS

-0.74

ernels

-0.72

ovies

-0.69

âĸº

-0.68

 slots

-0.67

eworks

-0.67

POSITIVE LOGITS

lier

0.71

 acknowledgment

0.70

 unto

0.68

hemer

0.68

letter

0.67

 homage

0.65

xiety

0.65

 antidote

0.64

thing

0.64

 testament

0.63

Activations Density 0.058%

adjectives followed by "of" and then a noun

phrases that categorize or describe types of entities or concepts

No Comments

No Known Activations