INDEX

Explanations

references to feelings of compulsion or inexplicable motivations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

for some reason

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Infórmanos

-0.48

portál

-0.46

 Numerade

-0.45

sumowanie

-0.45

Personensuche

-0.44

 gynhyrchwyd

-0.44

点此举报

-0.44

@"/

-0.44

 ویکی‌آمباردا

-0.44

価

-0.44

POSITIVE LOGITS

Somehow

0.65

 Somehow

0.64

 somehow

0.56

 weirdly

0.53

weird

0.53

 strangely

0.52

 weird

0.49

 mysteriously

0.49

 oddly

0.47

 Weird

0.45

Activations Density 0.014%

references to feelings of compulsion or inexplicable motivations

for some reason

No Comments

No Known Activations