INDEX

Explanations

inquiries about responsibility and knowledge surrounding incidents

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

ĸļ

-0.59

¥µ

-0.57

adle

-0.57

vice

-0.57

aughed

-0.55

elman

-0.54

Mobil

-0.54

Runner

-0.54

porate

-0.53

enegger

-0.53

POSITIVE LOGITS

 specifics

0.61

 significance

0.59

 whereabouts

0.58

 meanings

0.56

 motives

0.56

 exactly

0.56

nor

0.55

 exact

0.55

 passwords

0.55

channelAvailability

0.54

Activations Density 0.291%

inquiries about responsibility and knowledge surrounding incidents

No Comments

No Known Activations