INDEX

Explanations

sentences or phrases that contain significant questions or observations

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

start of explanatory clauses

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

 kasarigan

-0.86

 مرئيه

-0.81

 autorytatywna

-0.80

 ویکی‌آمباردا

-0.79

 للمعارف

-0.77

 queſta

-0.74

 ویکی‌پدیا

-0.73

viewDidLoad

-0.73

 صوتيه

-0.72

 Wiktionnaire

-0.72

POSITIVE LOGITS

but

0.33

0.31

<eos>

0.30

с

0.29

 pres

0.29

le

0.29

0.28

red

0.28

<0xAB>

0.28

Activations Density 0.097%

sentences or phrases that contain significant questions or observations

start of explanatory clauses

No Comments

No Known Activations