INDEX

Explanations

phrases related to political and legal terms

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

Nap

-0.86

VID

-0.67

 Hobby

-0.66

Franc

-0.65

Pa

-0.64

BIT

-0.62

iT

-0.62

Travel

-0.60

Moore

-0.60

Pak

-0.59

POSITIVE LOGITS

tml

0.84

orsi

0.81

phrine

0.78

clude

0.76

xual

0.76

igger

0.76

vernment

0.75

ary

0.70

allery

0.70

inkle

0.68

Activations Density 0.271%

phrases related to political and legal terms

No Comments

No Known Activations