INDEX

Explanations

programming and code structure elements like assignments and function calls

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

return [identifier name]

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 desierto

-0.37

 ordinaires

-0.31

 createDate

-0.30

 ujednoznacz

-0.30

seende

-0.30

 avanzada

-0.28

 noires

-0.28

 getPassword

-0.27

发表于

-0.26

 électroniques

-0.26

POSITIVE LOGITS

 ब्रेकडाउन

0.72

uxxxx

0.71

 noDo

0.67

 propOrder

0.65

ButterKnife

0.64

enterOuterAlt

0.64

sstream

0.63

synthesize

0.63

 arşivlendi

0.62



0.62

Activations Density 0.762%

programming and code structure elements like assignments and function calls

return [identifier name]

No Comments

No Known Activations