INDEX

Explanations

references to or mentions of locations, specifically New York City

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

backer

-0.78

wagen

-0.73

iasis

-0.71

ainer

-0.70

oise

-0.68

olate

-0.68

tenance

-0.68

 vain

-0.68

ial

-0.67

abase

-0.67

POSITIVE LOGITS

RB

1.21

SE

1.13

FW

1.07

NY

1.05

CHA

1.04

CHR

1.04

PD

1.01

RP

1.00

TC

0.99

WS

0.98

Activations Density 0.029%

references to or mentions of locations, specifically New York City

No Comments

No Known Activations