INDEX

Explanations

words and phrases that express concern and emotional impact

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

ãĥ¼ãĥĨãĤ£

-0.58

riched

-0.57

roots

-0.54

©¶æ

-0.53

 pione

-0.53

anqu

-0.52

ospital

-0.52

aughtered

-0.52

idal

-0.52

ideon

-0.51

POSITIVE LOGITS

:)

1.28

;)

1.25

 haha

1.20

 ðŁĻĤ

1.19

:-)

1.16

 ðŁĺ

1.03

lol

1.01

:(

1.01

tho

0.98

 anyways

0.93

Activations Density 0.520%

words and phrases that express concern and emotional impact

No Comments

No Known Activations