INDEX

Explanations

phrases indicating a specific type or category

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

 Halls

-0.60

stories

-0.60

edia

-0.59

 Rings

-0.57

 Bowman

-0.57

 Omaha

-0.57

 Minutes

-0.56

 sisters

-0.56

 Balls

-0.55

orney

-0.54

POSITIVE LOGITS

 imaginable

0.94

etting

0.84

 whatsoever

0.81

of

0.78

of

0.76

 thereof

0.76

icultural

0.74

lier

0.71

ĩ

0.68

stop

0.68

Activations Density 0.024%

phrases indicating a specific type or category

No Comments

No Known Activations