INDEX

Explanations

comparative phrases, especially those using 'as'

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

NEY

-0.63

ï¿½ï¿½ï¿½ï¿½

-0.62

DIT

-0.60

MAP

-0.59

ãģ®å®

-0.58

 respir

-0.56

MORE

-0.56

ETF

-0.55

 Examination

-0.55

bryce

-0.54

POSITIVE LOGITS

pired

1.39

pires

1.19

phy

1.07

ynchron

1.03

ym

1.03

pire

0.98

leep

0.97

far

0.95

imilar

0.95

 follows

0.94

Activations Density 0.303%

comparative phrases, especially those using 'as'

No Comments

No Known Activations