INDEX

Explanations

references to a specific individual's life events and achievements

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

itect

-0.74

daq

-0.74

needed

-0.71

ibaba

-0.71

hari

-0.71

ocument

-0.71

Trend

-0.70

inctions

-0.69

anguage

-0.69

å¥³

-0.69

POSITIVE LOGITS

own

1.69

 wife

1.37

 father

1.28

panic

1.28

 predecessor

1.27

 daughter

1.26

 successor

1.26

 Majesty

1.23

 nephew

1.23

 niece

1.22

Activations Density 1.172%

references to a specific individual's life events and achievements

No Comments

No Known Activations