INDEX

Explanations

phrases related to likelihood and probability

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

ASE

-0.69

 adamant

-0.67

 Edison

-0.62

ARCH

-0.62

 Episode

-0.61

athe

-0.59

 careful

-0.58

 lucky

-0.57

 praying

-0.57

Orn

-0.56

POSITIVE LOGITS

 exists

1.23

 revolves

1.18

 existed

1.18

 resembles

1.18

 belongs

1.18

 reflects

1.18

 occurs

1.17

 extends

1.14

 arose

1.14

 coincides

1.13

Activations Density 0.230%

phrases related to likelihood and probability

No Comments

No Known Activations