INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

istar

-0.70

gif

-0.65

yrinth

-0.64

icult

-0.63

between

-0.63

 nuts

-0.62

romy

-0.60

 guiActiveUnfocused

-0.59

nuts

-0.59

iths

-0.58

POSITIVE LOGITS

plate

1.44

plates

1.39

paces

1.06

akes

1.01

checked

0.99

calling

0.97

checks

0.91

 recognition

0.88

paced

0.84

ake

0.84

Activations Density 0.082%

proper nouns

references to names

No Comments

No Known Activations