INDEX

Explanations

non-English characters and numerical codes

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 brut

-0.84

 withd

-0.82

 manif

-0.79

 architectural

-0.74

 coping

-0.73

onga

-0.72

 metic

-0.72

 scaling

-0.71

 sustainability

-0.70

 inver

-0.70

POSITIVE LOGITS

 Retrieved

1.63

 Accessed

1.42

htm

1.23

 Includes

1.23

jpg

1.23

 Contains

1.22

 Available

1.21

html

1.18

png

1.17

txt

1.13

Activations Density 0.218%

non-English characters and numerical codes

No Comments

No Known Activations