INDEX

Explanations

phrases emphasizing the significance of roles or functions in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

arus

-0.68

False

-0.67

agle

-0.64

incinn

-0.63

atri

-0.63

Lev

-0.62

bones

-0.60

 jaws

-0.59

fts

-0.59

ilings

-0.58

POSITIVE LOGITS

 role

0.83

 helping

0.81

playing

0.79

 assisting

0.72

 facilitating

0.72

 mitigating

0.72

 roles

0.69

 supporting

0.68

 influencing

0.67

 toward

0.67

Activations Density 0.022%

phrases emphasizing the significance of roles or functions in various contexts

No Comments

No Known Activations