INDEX

Explanations

questions asking for opinions or perspectives on various topics

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

arm

-0.75

Ãį

-0.72

Thompson

-0.72

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.70

River

-0.70

Published

-0.70

aunder

-0.69

externalActionCode

-0.69

cycle

-0.69

reen

-0.68

POSITIVE LOGITS

...?

0.93

!?

0.81

 those

0.76

?!

0.73

?:

0.71

!?"

0.70

 fairness

0.70

 protecting

0.70

 grandchildren

0.68

Activations Density 0.019%

questions asking for opinions or perspectives on various topics

No Comments

No Known Activations