INDEX

Explanations

proper nouns, such as names of people, places, and organizations

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

draft

-0.63

otype

-0.62

cribed

-0.62

dozen

-0.61

erial

-0.61

orc

-0.61

higher

-0.58

rals

-0.58

eed

-0.57

perty

-0.57

POSITIVE LOGITS

 believes

0.80

 writes

0.77

 admits

0.75

 tells

0.72

 Says

0.69

 Productions

0.69

 wrote

0.68

 argues

0.68

 discusses

0.67

 enjoys

0.67

Activations Density 0.286%

proper nouns, such as names of people, places, and organizations

No Comments

No Known Activations