INDEX

Explanations

legal terms and information related to criminal defense and journalism

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

".[

-0.74

anqu

-0.66

'.

-0.63

ãĥł

-0.59

.[

-0.58

."[

-0.58

GN

-0.57

"-

-0.56

 Mund

-0.56

Us

-0.53

POSITIVE LOGITS

 please

1.00

 yourself

0.96

 PLEASE

0.88

 lately

0.83

or

0.78

?,

0.77

 blogging

0.76

 anymore

0.75

 podcasts

0.74

 yourselves

0.73

Activations Density 0.465%

legal terms and information related to criminal defense and journalism

No Comments

No Known Activations

legal terms and information related to criminal defense and journalism

No Comments

No Known Activations