INDEX

Explanations

references to document titles and section headers in markup or code

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

code tags and syntax

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

 pinulongan

-0.66

########.

-0.62

 وتسجيلات

-0.61

 informée

-0.58

RegressionTest

-0.58

脚注の使い方

-0.58

.",

-0.56

Diweddarwch

-0.55

 utafitiHapana

-0.54

."],

-0.54

POSITIVE LOGITS

 dieux

0.53

 invokingState

0.45

 güven

0.42

 mourir

0.40

 préfé

0.38

Due

0.38

 düş

0.38

due

0.38

 ış

0.38

 bezpiecze

0.37

Activations Density 0.005%

references to document titles and section headers in markup or code

code tags and syntax

No Comments

No Known Activations