INDEX

Explanations

proper nouns related to casting in films or television

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

vironment

-0.87

£ı

-0.83

¥µ

-0.77

ĸļ

-0.72

hops

-0.70

ĨĴ

-0.66

 Hipp

-0.65

 whichever

-0.64

pheus

-0.60

utherford

-0.59

POSITIVE LOGITS

aways

1.14

ration

1.05

member

1.00

lev

1.00

igated

0.95

members

0.95

rated

0.94

igating

0.92

rating

0.92

away

0.89

Activations Density 0.800%

proper nouns related to casting in films or television

No Comments

No Known Activations