INDEX

Explanations

names of geographical locations

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

 externalToEVAOnly

-0.73

aturdays

-0.71

nesday

-0.69

olphin

-0.66

ifference

-0.65

 Hours

-0.65

GF

-0.65

eryl

-0.65

aday

-0.64

Assembly

-0.64

POSITIVE LOGITS

¾

1.52

Į

1.46

Ĵ

1.41

©

1.39

Ķ

1.38

ĥ

1.36

¼

1.33

Ń·

1.30

ģ

1.29

¹

1.28

Activations Density 0.017%

names of geographical locations

No Comments

No Known Activations