INDEX

Explanations

comparative phrases that involve describing increasing or worsening intensity

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

¬¼

-0.84

pedia

-0.71

anka

-0.70

afety

-0.68

uta

-0.64

senal

-0.63

ctuary

-0.63

oyd

-0.61

acly

-0.60

ORPG

-0.60

POSITIVE LOGITS

than

1.59

 quicker

1.30

 clearer

1.29

 faster

1.26

 wiser

1.26

 richer

1.24

 stronger

1.23

 slower

1.22

 healthier

1.22

 than

1.22

Activations Density 0.280%

comparative phrases that involve describing increasing or worsening intensity

No Comments

No Known Activations