INDEX

Explanations

the word "David" in the text

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cffffcc

-1.01

NetMessage

-0.86

ĺħ

-0.80

phies

-0.76

Reply

-0.73

¥ŀ

-0.72

ĻĤ

-0.72

¥µ

-0.71

tnc

-0.70

*/(

-0.70

POSITIVE LOGITS

 Bowie

1.04

 Cameron

0.97

 Axel

0.91

 Petraeus

0.88

 Barton

0.87

son

0.87

 Rockefeller

0.87

 Beckham

0.86

Ic

0.84

 Koch

0.82

Activations Density 0.344%

the word "David" in the text

No Comments

No Known Activations