INDEX

Explanations

technical terms and references to coded language or commands

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

words and their actions

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

AndEndTag

-0.52

endphp

-0.46

 EconPapers

-0.45

OCCURRED

-0.44

ISupport

-0.41

 resourceCulture

-0.41

новниш

-0.40

GOTREF

-0.40

ragma

-0.40

 AssemblyTitle

-0.40

POSITIVE LOGITS

تقاوى

0.51

AsUp

0.47

 للمعارف

0.47

 initComponents

0.43

 jäsen

0.42

 între

0.38

 tartalomajánló

0.38

 dispoz

0.38

 rayures

0.38

 tyle

0.37

Activations Density 0.232%

technical terms and references to coded language or commands

words and their actions

No Comments

No Known Activations