INDEX

Explanations

instances of the word "stop" or its variations

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

Sov

-0.85

ammy

-0.77

ighth

-0.72

aths

-0.70

uries

-0.68

ramid

-0.67

ocene

-0.67

arov

-0.67

orthy

-0.65

é¾įåĸļå£«

-0.65

POSITIVE LOGITS

 bothering

1.07

gap

0.98

 raining

0.90

watch

0.89

watching

0.88

 breathing

0.87

 smoking

0.85

 worrying

0.83

 abruptly

0.83

 bleeding

0.78

Activations Density 0.381%

instances of the word "stop" or its variations

No Comments

No Known Activations