INDEX

Explanations

medical and technological terms and concepts

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

ãĥ´ãĤ¡

-0.80

ãĥ´

-0.71

Ļ½

-0.70

ãĥĦ

-0.66

soType

-0.65

oho

-0.64

wcs

-0.63

spo

-0.63

ãĥ¥

-0.62

:(

-0.61

POSITIVE LOGITS

/"

0.69

 designation

0.66

ventures

0.65

psey

0.64

 mode

0.63

 hybrid

0.61

 membranes

0.59

 Complex

0.58

 hybrids

0.58

 moniker

0.57

Activations Density 0.097%

medical and technological terms and concepts

No Comments

No Known Activations