INDEX

Explanations

phrases related to scientific measurements and calibration

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

reference standard

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

AddTagHelper

-0.38

lateinit

-0.38

huur

-0.38

instancetype

-0.35

handleSubmit

-0.35

 فريبيس

-0.35

 partea

-0.34

cokinetics

-0.34

VersionUID

-0.34

 باسم

-0.34

POSITIVE LOGITS

tonode

0.54

tonsoft

0.43

ayak

0.41

Reference

0.40

 MutableLiveData

0.39

nsic

0.39

Havolalar

0.38

 gynhyrchwyd

0.37

 reference

0.37

Chham

0.36

Activations Density 0.169%

phrases related to scientific measurements and calibration

reference standard

No Comments

No Known Activations