INDEX

Explanations

indicators of political and economic critique or analysis

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

bye

-0.67

 Himself

-0.64

 americ

-0.62

 vert

-0.60

 microwave

-0.59

#$

-0.57

 fulfil

-0.57

 kinda

-0.56

OK

-0.56

 merge

-0.56

POSITIVE LOGITS

ccording

0.85

ebus

0.80

0.75

,.

0.71

irteen

0.67

acket

0.64

 though

0.63

adays

0.63

itia

0.62

ohn

0.61

Activations Density 0.166%

indicators of political and economic critique or analysis

No Comments

No Known Activations