INDEX

Explanations

code-related syntax and specific programming constructs

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

Code, includes, or user prompts

np_acts-logits-general · gemini-2.0-flash

code tokens 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 @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

parsedMessage

-0.73

awtextra

-0.70

 surla

-0.68

 ujednoznacz

-0.66

 pinulongan

-0.66

 незавершена

-0.65

]")]

-0.63

Tikang

-0.59

-------

-0.59

)‏

-0.57

POSITIVE LOGITS

Linq

0.46

 vandens

0.38

 OnInit

0.38

 šal

0.36

 neón

0.33

 inoxidable

0.33

 возможности

0.33

 atlik

0.33

 cruce

0.32

 spørs

0.32

Activations Density 0.012%

code-related syntax and specific programming constructs

Code, includes, or user prompts

code tokens and special characters

No Comments

No Known Activations