INDEX

Explanations

proper nouns related to names and characters in various contexts

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

CPC

-0.63

mercial

-0.61

 costing

-0.60

 suburb

-0.60

 IMAGES

-0.58

 UNESCO

-0.57

 staffed

-0.57

 subsistence

-0.57

 federally

-0.56

 childcare

-0.56

POSITIVE LOGITS

isha

0.79

nar

0.78

ira

0.76

orius

0.76

anka

0.76

yss

0.74

ayne

0.73

yson

0.73

yk

0.72

rius

0.72

Activations Density 0.443%

proper nouns related to names and characters in various contexts

No Comments

No Known Activations