INDEX

Explanations

medical terminology and health-related information

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

Ò

-0.74

CHAPTER

-0.69

thood

-0.69

leground

-0.68

abetes

-0.68

leeve

-0.68

icia

-0.68

!!

-0.67

NB

-0.66

!!!!

-0.65

POSITIVE LOGITS

 slightest

1.47

 entire

1.30

 majority

1.21

 latter

1.16

 remainder

1.14

 whole

1.13

 vast

1.10

 difference

1.09

 easiest

1.08

oret

1.07

Activations Density 0.440%

medical terminology and health-related information

No Comments

No Known Activations