INDEX

Explanations

dates and events mentioned in news articles

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

fried

-0.59

 understatement

-0.58

Others

-0.58

OME

-0.58

ĺħ

-0.57

²¾

-0.56

 predic

-0.56

(-

-0.54

entimes

-0.53

finder

-0.53

POSITIVE LOGITS

PDT

0.78

EDT

0.71

PST

0.68

UTC

0.66

when

0.65

 Mehran

0.65

 Accessed

0.64

GMT

0.64

 wherein

0.63

 pione

0.62

Activations Density 0.113%

dates and events mentioned in news articles

No Comments

No Known Activations