INDEX

Explanations

phrases indicating uncertainty or negative outcomes

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

foreign language punctuation and conjunctions

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

enters

-0.34

 enige

-0.33

 تضيفلها

-0.31

月号

-0.30

codable

-0.30

delwed

-0.30

хьтан

-0.30

Bibliograf

-0.29

喱

-0.29

property

-0.28

POSITIVE LOGITS

rungsseite

0.66

MLLoader

0.60

__::

0.56

دانشنامهٔ

0.54

 informée

0.53

yntaxException

0.52

Ӕ

0.51

protoimpl

0.50



0.49

uxxxx

0.49

Activations Density 0.065%

phrases indicating uncertainty or negative outcomes

foreign language punctuation and conjunctions

No Comments

No Known Activations