INDEX

Explanations

specific programming language syntax or function calls

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

programming and data

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

 noDo

-0.70

expandindo

-0.67

 ujednoznacz

-0.65

transQ

-0.65

 keyCode

-0.60

::_('

-0.58

writeFieldEnd

-0.58

 queſta

-0.56

󠁴

-0.55

 NgModule

-0.54

POSITIVE LOGITS

Slf

0.38

awtextra

0.35

 صوتيه

0.32

笑了笑

0.32

 unnoticed

0.31

kuun

0.30

 Very

0.29

Hidden

0.28

 Rather

0.28

fördert

0.28

Activations Density 0.002%

specific programming language syntax or function calls

programming and data

No Comments

No Known Activations