INDEX

Explanations

descriptors related to appearances and finishes, particularly those that convey a sense of shine or polish

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

orders

-0.79

ffic

-0.73

zee

-0.71

acl

-0.71

ommod

-0.68

agents

-0.68

MQ

-0.68

argument

-0.67

population

-0.66

WER

-0.65

POSITIVE LOGITS

 stainless

0.87

 sterling

0.82

 rusty

0.81

 finish

0.81

 steel

0.80

 teeth

0.79

 polished

0.79

 slick

0.77

 chrome

0.77

 polish

0.76

Activations Density 0.067%

descriptors related to appearances and finishes, particularly those that convey a sense of shine or polish

No Comments

No Known Activations