INDEX

Explanations

words indicative of significant events or releases

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

dates or numbers

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

دانشنامهٔ

-0.49

 Chwiliwch

-0.46

angliski

-0.42

afficheront

-0.40

ValueStyle

-0.40

 resourceCulture

-0.39

 kasa

-0.38

𝟹

-0.36

المراجع

-0.36

 bílá

-0.36

POSITIVE LOGITS

IVEREF

0.51

marion

0.49

 ambigu

0.47

RTEE

0.46

 Suga

0.45

AccessorTable

0.44

mello

0.44

introduced

0.44

idea

0.44

ofition

0.44

Activations Density 0.023%

words indicative of significant events or releases

dates or numbers

No Comments

No Known Activations