INDEX

Explanations

words related to desire or willingness

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

issance

-0.87

ework

-0.69

bard

-0.69

 Enhancement

-0.69

unny

-0.67

icol

-0.66

igmatic

-0.65

ilian

-0.65

Kings

-0.64

utical

-0.63

POSITIVE LOGITS

 anymore

0.94

 anybody

0.94

 anything

0.93

 anyone

0.91

nor

0.83

any

0.83

ANY

0.78

ANY

0.77

 them

0.74

 undue

0.72

Activations Density 0.039%

words related to desire or willingness

No Comments

No Known Activations