INDEX

Explanations

dialogue markers used in a conversation

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

 Mechdragon

-0.73

SourceFile

-0.73

alion

-0.71

tesy

-0.70

 Citiz

-0.67

à¤¾

-0.67

itionally

-0.67

iership

-0.66

eday

-0.66

ochet

-0.66

POSITIVE LOGITS

uh

1.28

um

1.05

yeah

0.97

 yeah

0.96

maybe

0.86

sort

0.86

 gasp

0.85

uh

0.84

wow

0.82

wait

0.82

Activations Density 0.074%

dialogue markers used in a conversation

No Comments

No Known Activations