INDEX

Explanations

various forms of the word "become" or "became."

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ength

-0.73

ramid

-0.70

ongyang

-0.70

inately

-0.67

 goodness

-0.67

erity

-0.67

inarily

-0.66

idth

-0.65

iology

-0.64

bell

-0.63

POSITIVE LOGITS

 extinct

1.08

 accustomed

1.01

 entangled

1.01

 embroiled

0.95

 acquainted

0.94

 obsolete

0.90

 aware

0.87

 synonymous

0.86

 obsessed

0.84

 increasingly

0.84

Activations Density 0.102%

various forms of the word "become" or "became."

No Comments

No Known Activations