INDEX

Explanations

adverbs describing degree or amount

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

agate

-0.73

ourses

-0.73

alis

-0.72

 City

-0.69

Ds

-0.67

city

-0.66

endi

-0.66

City

-0.66

ogi

-0.65

oran

-0.65

POSITIVE LOGITS

 identical

0.94

 nonexistent

0.89

 impossible

0.88

 indistinguishable

0.87

 unanimous

0.85

 useless

0.83

 extinct

0.82

zero

0.79

 worthless

0.77

 meaningless

0.76

Activations Density 2.609%

adverbs describing degree or amount

No Comments

No Known Activations