INDEX

Explanations

references to personal anecdotes and experiences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 guiName

-0.70

tesy

-0.68

ãĥ«

-0.68

manac

-0.67

":"/

-0.63

obook

-0.63

itled

-0.61

afe

-0.61

theless

-0.60

ublished

-0.60

POSITIVE LOGITS

etc

2.00

etc

1.66

whatever

1.21

 blah

1.20

ect

1.11

...)

1.06

âĢ¦)

1.05

 whatever

1.01

et

0.85

yes

0.80

Activations Density 0.185%

references to personal anecdotes and experiences

No Comments

No Known Activations