INDEX

Explanations

references to page numbers and editorial content

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

story page 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

 Administrativna

-0.53

okovic

-0.37

 portanto

-0.35

 ब्रेकडाउन

-0.35

urança

-0.34

厉害

-0.34

 tuttavia

-0.34

 undangan

-0.34

 zodia

-0.34

 therefore

-0.33

POSITIVE LOGITS

stdc

0.54

awaiter

0.53

 doInBackground

0.53

期刊论文

0.50

dotenv

0.49

 MainAxisSize

0.49

autogui

0.49

PerformLayout

0.49

AutoScale

0.49

qxd

0.48

Activations Density 0.011%

references to page numbers and editorial content

story page numbers

No Comments

No Known Activations