INDEX

Explanations

sequences of underscores or blank spaces, indicating missing or redacted information

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

P.3d ___

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

<bos>

-0.97

liceerd

-0.65

__(/*!

-0.62



-0.60

interopRequire

-0.59

 المعيارى

-0.59

jsonwebtoken

-0.57

serviceWorker

-0.56

Juven

-0.56

marvin

-0.56

POSITIVE LOGITS

___

1.13

___

0.65

0.59

setBounds

0.47

(___

0.46

____

0.46

sweise

0.46

>-->

0.43

⠀

0.42

ecka

0.39

Activations Density 0.046%

sequences of underscores or blank spaces, indicating missing or redacted information

P.3d ___

No Comments

No Known Activations