INDEX

Explanations

information regarding laws and governmental actions

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

case

-0.76

Untitled

-0.70

azo

-0.68

iots

-0.67

OK

-0.66

pee

-0.66

Fuck

-0.66

Shut

-0.65

urg

-0.64

sex

-0.64

POSITIVE LOGITS

 consequently

1.25

 therefore

1.25

 hence

1.23

 moreover

1.15

 although

1.15

 thus

1.12

 furthermore

1.10

 doubtless

1.03

 luckily

1.01

 fortunately

0.98

Activations Density 1.298%

information regarding laws and governmental actions

No Comments

No Known Activations