INDEX

Explanations

references to coded language or slang

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

technical language and slang

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

 central

-0.38

ByPrimaryKey

-0.29

 confronti

-0.29

SpringBoot

-0.29

 basan

-0.28

Clik

-0.28

раздо

-0.28

 explicitly

-0.28

 orsz

-0.27

 histórica

-0.27

POSITIVE LOGITS

tagHelperRunner

0.74

 slang

0.67

 jargon

0.66

AndEndTag

0.65

 langage

0.63

 propOrder

0.62

httphttps

0.62

webElementXpaths

0.60

 uLocal

0.59

 uParam

0.59

Activations Density 0.125%

references to coded language or slang

technical language and slang

No Comments

No Known Activations