INDEX

Explanations

the symbol "<" typically used in programming or markup languages

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

API parsing and data structures

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

Negative Logits

TabIndex

-0.45

 resourceCulture

-0.38

 hinzu

-0.35

UrlResolution

-0.34

一愣

-0.33

 mosa

-0.32

 HasFactory

-0.32

endmodule

-0.32

AddTagHelper

-0.31

Getenv

-0.30

POSITIVE LOGITS

:✨

0.61

elemField

0.57

parsedMessage

0.55

цездатний

0.52

 defaultstate

0.51

 للمعارف

0.50

gnostic

0.50

 незавершена

0.50

tvguidetime

0.49

 vermelhas

0.48

Activations Density 0.016%

the symbol "<" typically used in programming or markup languages

API parsing and data structures

No Known Activations