INDEX

Explanations

phrases indicating certainty or agreement

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

foreseen

-0.74

natureconservancy

-0.69

tnc

-0.68

Vert

-0.65

mone

-0.65

 resil

-0.62

andom

-0.62

govtrack

-0.61

encer

-0.61

MpServer

-0.61

POSITIVE LOGITS

ndra

0.99

 enough

0.94

entimes

0.89

terday

0.83

footed

0.76

ty

0.74

ly

0.67

 Gamble

0.64

ties

0.63

ness

0.62

Activations Density 0.013%

phrases indicating certainty or agreement

No Comments

No Known Activations