INDEX

Explanations

phrases related to critique or evaluation

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fml

-0.81

NetMessage

-0.75

/"

-0.73

wcsstore

-0.71

>(

-0.71

Catalog

-0.70

Edge

-0.68

 chops

-0.68

................................................................

-0.68

BILITIES

-0.67

POSITIVE LOGITS

 mankind

0.86

 humankind

0.81

 those

0.75

our

0.72

 humanity

0.72

 civilization

0.72

 tomorrow

0.70

 ordinary

0.69

 civilisation

0.69

 academia

0.68

Activations Density 1.043%

phrases related to critique or evaluation

No Comments

No Known Activations