INDEX

Explanations

names and features of various applications or tools

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

subject completing phrases

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

 nahilalakip

-0.81

tagHelperRunner

-0.70

exitRule

-0.69

 protoimpl

-0.63

 nakalista

-0.61

TypedDataSet

-0.58

참고

-0.56

 ویکی‌پدی

-0.56

 esternos

-0.55

клопе

-0.52

POSITIVE LOGITS

 gants

0.33

 zeker

0.30

CTL

0.28

ще

0.28

 ка

0.28

 regels

0.28

occasione

0.28

vay

0.28

 ladder

0.27

CCR

0.27

Activations Density 0.022%

names and features of various applications or tools

subject completing phrases

No Comments

No Known Activations