INDEX

Explanations

phrases related to significant or impactful events, actions, or entities

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

Dialogue

-0.70

 Emin

-0.69

cia

-0.68

istry

-0.67

idency

-0.66

yrim

-0.66

ILA

-0.66

DragonMagazine

-0.65

 Reincarnated

-0.64

manship

-0.63

POSITIVE LOGITS

oted

1.27

gest

1.21

gie

1.07

wig

1.02

 chunk

0.96

ol

0.96

 bang

0.96

ger

0.95

GER

0.95

 bucks

0.90

Activations Density 0.311%

phrases related to significant or impactful events, actions, or entities

No Comments

No Known Activations