INDEX

Explanations

references to live performances or shows

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

 calcul

-0.70

tymology

-0.69

 notor

-0.66

 solitary

-0.64

etheless

-0.61

srf

-0.59

 captcha

-0.58

 nutshell

-0.57

 saline

-0.56

 Combine

-0.56

POSITIVE LOGITS

biz

1.50

manship

1.38

rooms

1.30

room

1.28

downs

1.22

case

1.20

alter

1.13

cases

1.09

boat

1.06

runners

1.04

Activations Density 0.029%

references to live performances or shows

No Comments

No Known Activations