INDEX

Explanations

personal pronouns and references to unspecified entities

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

idth

-0.63

 Frontier

-0.62

ILE

-0.62

 Pirates

-0.58

ãĥ©ãĥ³

-0.58

 Passenger

-0.57

 Friend

-0.56

 Balloon

-0.56

 Wanted

-0.56

 Petroleum

-0.55

POSITIVE LOGITS

alian

1.39

self

1.26

chy

1.17

unes

1.06

selves

1.01

atic

1.00

atically

0.96

iner

0.91

MpServer

0.85

ueller

0.80

Activations Density 1.161%

personal pronouns and references to unspecified entities

No Comments

No Known Activations