INDEX

Explanations

words related to influence, regulation, and impact

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

mad

-0.75

DragonMagazine

-0.73

iphate

-0.69

ãĤ»

-0.69

 sidx

-0.67

rit

-0.65

 topped

-0.65

shake

-0.65

hey

-0.63

vic

-0.63

POSITIVE LOGITS

ively

0.95

 them

0.92

orously

0.87

our

0.83

uate

0.80

the

0.80

 aspects

0.78

him

0.77

ibly

0.76

ably

0.72

Activations Density 1.725%

words related to influence, regulation, and impact

No Comments

No Known Activations