INDEX

Explanations

references to diseases and outbreaks

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ramid

-0.78

racuse

-0.72

CHQ

-0.70

 Idol

-0.69

eele

-0.69

AZ

-0.68

robe

-0.67

cise

-0.66

assy

-0.66

icrobial

-0.65

POSITIVE LOGITS

 plag

1.55

 caused

1.30

 plague

1.21

 outbreak

1.19

 epidemic

1.18

 prevention

1.16

 Symptoms

1.14

 severity

1.12

 worsened

1.11

 symptoms

1.09

Activations Density 3.128%

references to diseases and outbreaks

No Comments

No Known Activations