INDEX

Explanations

references to physical confrontation and violence

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

Quantity

-0.60

 humankind

-0.57

 Explos

-0.56

 citizens

-0.55

(%)

-0.55

 swallowed

-0.55

 Atkinson

-0.55

ario

-0.54

 WRITE

-0.53

 Laksh

-0.53

POSITIVE LOGITS

 whom

1.19

 affection

0.78

who

0.77

brother

0.76

legram

0.75

igslist

0.74

oba

0.71

assad

0.67

 pseudonym

0.65

illac

0.64

Activations Density 3.038%

references to physical confrontation and violence

No Comments

No Known Activations