INDEX

Explanations

specific programming and data structure-related terms and constructs

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

code constructs and special characters

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

 Todes

-0.32

es

-0.31

 "..\..\

-0.30

 "..\..\..\

-0.30

 denn

-0.30

letra

-0.29

 inte

-0.29

siune

-0.28

tow

-0.28

案内

-0.28

POSITIVE LOGITS

 Савезне

0.74

Tikang

0.69

 OMITBAD

0.68

oa̍t

0.67

 estekak

0.67

 للاسماء

0.67

IndentedString

0.66

 nahilalakip

0.65

expandindo

0.64

 Paglinawan

0.63

Activations Density 0.002%

specific programming and data structure-related terms and constructs

code constructs and special characters

No Comments

No Known Activations