INDEX

Explanations

metadata annotations and serialized object descriptors

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

code and non-English 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

AddHtmlAttribute

-0.66

 betweenstory

-0.61

aarrggbb

-0.61

MessageTagHelper

-0.58

 utafitiHapana

-0.58

ConstraintMaker

-0.57

stdc

-0.56

RectangleBorder

-0.56

 BorderSide

-0.56

disposing

-0.56

POSITIVE LOGITS

يكب

0.29

లాలు

0.29

AE

0.26

 tica

0.25

 aidé

0.25

ांकि

0.25

nextDouble

0.25

 обще

0.24

(({

0.24

ay

0.24

Activations Density 0.008%

metadata annotations and serialized object descriptors

code and non-English words

No Comments

No Known Activations