INDEX

Explanations

words related to the act of shortening or reducing

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

rera

-0.67

 unic

-0.67

rero

-0.64

 Chaff

-0.63

PAR

-0.63

ublic

-0.63

ashington

-0.60

cffffcc

-0.59

dinand

-0.58

è¦ļéĨĴ

-0.58

POSITIVE LOGITS

ened

0.79

ners

0.71

imus

0.71

forth

0.70

emies

0.70

 Flavoring

0.69

heartedly

0.69

etr

0.66

mble

0.65

ning

0.65

Activations Density 0.026%

words related to the act of shortening or reducing

No Comments

No Known Activations