INDEX

Explanations

organizations, specific numbers, and business-related 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_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

urgy

-0.68

 takeoff

-0.66

 havoc

-0.62

inction

-0.62

ciation

-0.62

lihood

-0.61

ageddon

-0.60

 Syndrome

-0.60

ipation

-0.59

chwitz

-0.58

POSITIVE LOGITS

 colleagues

1.19

 fellow

1.17

 onlook

1.13

 strangers

1.07

 classmates

1.06

 friends

1.06

 professionals

1.03

 reviewers

1.02

 teammates

1.01

 colleague

1.01

Activations Density 2.522%

organizations, specific numbers, and business-related terms

No Comments

No Known Activations