INDEX

Explanations

Twitter usernames and related information about social media interactions

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

prints

-0.67

backer

-0.62

 Turing

-0.58

 relational

-0.58

©¶æ

-0.58

 phenomen

-0.57

ters

-0.57

 flares

-0.56

 polymorph

-0.56

sofar

-0.55

POSITIVE LOGITS

irus

1.28

oided

1.15

oice

1.09

arro

1.08

enture

1.07

ocado

1.06

ille

1.03

iew

1.02

irtual

1.02

ascular

1.02

Activations Density 0.021%

Twitter usernames and related information about social media interactions

No Comments

No Known Activations