INDEX

Explanations

words related to legal cases and court rulings

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

 Cinderella

-0.63

ãĥ£

-0.62

 Mermaid

-0.60

qt

-0.60

 Garland

-0.59

ãĢĲ

-0.57

 Cork

-0.56

reau

-0.56

iard

-0.56

 Reborn

-0.55

POSITIVE LOGITS

igent

0.92

aneous

0.80

elta

0.78

ACTION

0.78

lycer

0.78

urnal

0.76

entary

0.75

nel

0.73

olated

0.73

ensive

0.71

Activations Density 0.036%

words related to legal cases and court rulings

No Comments

No Known Activations