INDEX

Explanations

specific and technical identifiers or symbols related to coding and programming

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

acr credential

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

 ModelExpression

-0.50

########.

-0.43

ArrowToggle

-0.39

ceğine

-0.37

uxxxx

-0.35

 channels

-0.35

lavi

-0.33

potent

-0.33

wagon

-0.33

 /*#__

-0.33

POSITIVE LOGITS

 الحره

0.51

 שוליים

0.51

acidade

0.50

 Exactos

0.50

Datuak

0.49

Fordítás

0.49

Răspunde

0.44

 esprits

0.44

참고

0.43

Manbalar

0.43

Activations Density 0.000%

specific and technical identifiers or symbols related to coding and programming

acr credential

No Comments

No Known Activations