INDEX

Explanations

statements expressing varying levels of certainty and opinions

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

sac

-0.73

 Delicious

-0.69

 Alone

-0.68

 Himself

-0.67

 Maintenance

-0.64

HIS

-0.63

ãĥĺ

-0.63

 Shutterstock

-0.62

ZIP

-0.61

Fug

-0.61

POSITIVE LOGITS

 speculate

1.22

 speculated

1.02

 commentators

0.98

 accuse

0.96

 disbel

0.95

 clam

0.95

 argue

0.91

 ridicule

0.90

 skept

0.88

 questioning

0.85

Activations Density 2.184%

statements expressing varying levels of certainty and opinions

No Comments

No Known Activations