INDEX

Explanations

references to social dynamics and group interactions

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

contextually descriptive words

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

 ویکی‌پدی

-0.35

 Biôgrafia

-0.33

 Walkover

-0.32

 cosi

-0.31

GEM

-0.31

 ErrIntOverflow

-0.30

Domain

-0.30

anément

-0.30

Nasty

-0.29

reportWebVitals

-0.29

POSITIVE LOGITS

///</

0.63

twimg

0.60

__(/*!

0.59



0.58

sizeCache

0.53

 endblock

0.51

XmlAccessorType

0.50

tvguidetime

0.50

 <>",

0.50

jsxFileName

0.48

Activations Density 0.153%

references to social dynamics and group interactions

contextually descriptive words

No Comments

No Known Activations