INDEX

Explanations

references to size, magnitude, or quantity, especially with a focus on comparison or measurement

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

esson

-0.83

vous

-0.80

eering

-0.76

odge

-0.73

odore

-0.71

ĪĴ

-0.70

pron

-0.69

ournal

-0.68

 Leilan

-0.68

swick

-0.67

POSITIVE LOGITS

 scale

0.95

 scales

0.81

 replica

0.81

 invari

0.78

 scaled

0.75

 proportions

0.75

Factor

0.75

Scale

0.74

 fabrication

0.71

 cooker

0.71

Activations Density 0.020%

references to size, magnitude, or quantity, especially with a focus on comparison or measurement

No Comments

No Known Activations