INDEX

Explanations

terms related to legal or law enforcement procedures and corporate business deals

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

HUD

-0.67

iscons

-0.66

oday

-0.65

iyah

-0.64

aye

-0.62

utsch

-0.62

iHUD

-0.61

isen

-0.61

LET

-0.60

çīĪ

-0.60

POSITIVE LOGITS

 nowadays

1.07

 rarely

0.88

 dictators

0.86

 normally

0.85

 typically

0.82

 seldom

0.82

 notoriously

0.81

 Usually

0.81

 usually

0.81

 tends

0.80

Activations Density 1.432%

terms related to legal or law enforcement procedures and corporate business deals

No Comments

No Known Activations