INDEX

Explanations

quotations that are the speaker's statements

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

etheless

-1.01

ingu

-0.77

£ı

-0.73

¬¼

-0.70

anuts

-0.66

»Ĵ

-0.66

Ĥª

-0.64

bably

-0.64

²¾

-0.62

osite

-0.62

POSITIVE LOGITS

 said

1.23

said

1.22

 says

1.21

 reads

1.18

 wrote

1.10

 commented

1.03

 recalls

1.00

 writes

0.98

 joked

0.97

 remarked

0.97

Activations Density 0.390%

quotations that are the speaker's statements

No Comments

No Known Activations