INDEX

Explanations

phrases related to processes and actions involving systems and design

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

properly design and calibrate

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

脚注の使い方

-0.72

 SwitchCompat

-0.68

Бахар

-0.64

expandindo

-0.60

 referenties

-0.60

 snippetHide

-0.60

Попис

-0.59

contentLoaded

-0.57

 Chwiliwch

-0.54

énario

-0.53

POSITIVE LOGITS

GetAsync

0.33

 taş

0.31

 vetorial

0.30

setOptions

0.30

 sagrado

0.30

 religieuses

0.30

 transición

0.30

 computadora

0.30

 magnético

0.29

 vettoriale

0.28

Activations Density 0.148%

phrases related to processes and actions involving systems and design

properly design and calibrate

No Comments

No Known Activations