INDEX

Explanations

people and direct them to take some action or support a cause

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

 Shap

-0.67

ongh

-0.64

ixties

-0.64

TED

-0.63

terday

-0.62

onday

-0.60

olate

-0.60

angular

-0.59

inventoryQuantity

-0.58

 Neuroscience

-0.57

POSITIVE LOGITS

to

0.76

 permission

0.72

 reinvest

0.68

DERR

0.67

 reconsider

0.65

 intervention

0.64

 join

0.64

 sake

0.64

 forgiveness

0.64

 secretaries

0.63

Activations Density 1.545%

people and direct them to take some action or support a cause

No Comments

No Known Activations