INDEX

Explanations

details related to medical treatment or interventions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rooft

-0.75

 RAID

-0.73

 dissenting

-0.71

 quotes

-0.70

 Franch

-0.67

 classy

-0.67

ãĥķãĤ©

-0.67

 hasht

-0.66

 slogans

-0.66

 quote

-0.66

POSITIVE LOGITS

 adulthood

1.60

 onset

1.40

 puberty

1.35

 adolescence

1.30

uberty

1.29

 maturity

1.17

age

1.13

regnancy

1.11

 infancy

1.09

 periods

1.09

Activations Density 0.285%

details related to medical treatment or interventions

No Comments

No Known Activations