INDEX

Explanations

phrases related to bonuses, numbers, and specific recommendations

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

door switch

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

uxxxx

-0.45

탤

-0.38

rungsseite

-0.38

polation

-0.37

 demonios

-0.36

AddTagHelper

-0.36

Citiți

-0.36

 själva

-0.35

bibitem

-0.35

TokenNameLPAREN

-0.35

POSITIVE LOGITS

ScopeManager

0.58

webElementXpaths

0.57

ᅠ

0.56

Firestore

0.50

AMOS

0.50

 متعلقه

0.46

 Ceram

0.46

MEG

0.45

InstanceState

0.45

انجليز

0.45

Activations Density 0.008%

phrases related to bonuses, numbers, and specific recommendations

door switch

No Comments

No Known Activations