INDEX

Explanations

references to blog posts and blog-related content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.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

 Lauder

-0.81

 evils

-0.66

BuyableInstoreAndOnline

-0.65

odka

-0.64

 Plat

-0.60

 Fifa

-0.60

 Strauss

-0.60

iod

-0.60

 crumbling

-0.59

 condoms

-0.59

POSITIVE LOGITS

gers

1.28

osphere

1.19

ger

1.10

ged

1.10

ging

1.10

spot

1.05

lov

1.02

post

0.94

bub

0.94

gres

0.89

Activations Density 0.699%

references to blog posts and blog-related content

No Comments

No Known Activations