INDEX

Explanations

loved ones or expressions of pride and heroism related to them

oai_token-act-pair · gpt-3.5-turbo 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulated

-0.73

ulhu

-0.70

uration

-0.66

é¾

-0.65

ulation

-0.65

UL

-0.65

 Regulatory

-0.65

IDER

-0.65

ilion

-0.64

ural

-0.63

POSITIVE LOGITS

 dearly

1.01

 uncond

0.92

 pets

0.86

liest

0.84

 loved

0.84

 nephew

0.81

 ones

0.81

 aunt

0.80

 grandchildren

0.80

 niece

0.78

Activations Density 0.033%

loved ones or expressions of pride and heroism related to them

No Comments

No Known Activations