INDEX

Explanations

frequency of phrases indicating numerical ranking or assessments

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ynes

-0.67

wa

-0.66

quel

-0.64

iko

-0.64

Reviewer

-0.63

iture

-0.62

WARE

-0.62

nod

-0.61

aeus

-0.61

 Bought

-0.59

POSITIVE LOGITS

 civilization

0.79

 humanity

0.74

 humankind

0.73

 civilisation

0.71

 mankind

0.69

 hostilities

0.68

 sorts

0.68

our

0.68

 what

0.67

 adulthood

0.67

Activations Density 0.102%

frequency of phrases indicating numerical ranking or assessments

No Comments

No Known Activations