INDEX

Explanations

the word "Sam" in texts

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

err

-0.66

 fences

-0.62

 GDDR

-0.62

 troll

-0.61

cffff

-0.60

 trolling

-0.59

_-

-0.58

 diving

-0.56

 busiest

-0.55

 CONTIN

-0.55

POSITIVE LOGITS

ples

1.29

pling

1.29

urai

1.27

pler

1.16

oan

1.16

uel

1.03

pled

1.00

mons

1.00

mys

0.95

ara

0.94

Activations Density 0.011%

the word "Sam" in texts

No Comments

No Known Activations