INDEX

Explanations

statements related to legislation and political discussions

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

Bos

-0.71

 Romans

-0.69

 Freak

-0.69

 Crusader

-0.66

 Hats

-0.63

 Salam

-0.63

 Pony

-0.62

 Towers

-0.62

 Spears

-0.62

 Dungeons

-0.61

POSITIVE LOGITS

rogens

0.90

rogen

0.84

 rehabilit

0.83

 dissemin

0.79

 distribute

0.78

 refine

0.77

 manipulate

0.77

/+

0.77

 circulate

0.76

 analyze

0.76

Activations Density 1.856%

statements related to legislation and political discussions

No Comments

No Known Activations