INDEX

Explanations

information related to locations and public access

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

olate

-0.72

doi

-0.70

 Canaver

-0.69

abase

-0.65

olin

-0.65

terness

-0.64

wark

-0.62

lihood

-0.61

 Scotia

-0.60

DIT

-0.60

POSITIVE LOGITS

 tourists

1.21

 foreigners

1.11

 visitors

1.07

 pilgrims

1.01

 refugees

1.00

 guests

0.97

 overflow

0.97

 occupants

0.96

 migrants

0.96

 immigrants

0.95

Activations Density 2.258%

information related to locations and public access

No Comments

No Known Activations