INDEX

Explanations

legal terms and court-related language

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

ku

-0.73

 Saban

-0.68

ulic

-0.67

zsche

-0.67

yip

-0.67

eco

-0.64

edia

-0.64

ec

-0.63

Ã´

-0.61

ombies

-0.61

POSITIVE LOGITS

 judge

1.13

 Judge

1.08

rooms

0.98

 magistrate

0.97

istrate

0.91

Court

0.90

 Settlement

0.90

 Clerk

0.89

Judge

0.89

 proceedings

0.87

Activations Density 0.036%

legal terms and court-related language

No Comments

No Known Activations