Skip to content
TheChatArchiveThe Chat Archive
Kimikimi-k2-0711-preview· Moonshot AIHugging Face LMArena Human Preference 140K

\begin{figure}[ht] \centering \begin{tikzpicture}[ process/.style={rectangle, draw,…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Metallama-4-scout-17b-16e-instruct’s answer.
  1. User

    \begin{figure}[ht] \centering \begin{tikzpicture}[ process/.style={rectangle, draw, minimum height=0.7cm, minimum width=2.2cm, align=center, fill=yellow!10, font=\footnotesize}, decision/.style={diamond, draw, minimum height=0.7cm, minimum width=1.8cm, align=center, fill=red!10, font=\footnotesize}, startend/.style={ellipse, draw, minimum height=0.7cm, minimum width=1.8cm, align=center, fill=green!10, font=\footnotesize}, arrow/.style={-Stealth, thick}, node distance=1.2cm and 0.8cm, scale=0.85, transform shape ]

    % Nodes \node[startend] (start) {Start}; \node[process, below=of start] (sensors) {Read Sensors}; \node[process, below=of sensors] (json) {Create JSON}; \node[process, below=of json] (serial) {Send to ESP32}; \node[decision, below=of serial] (wifi) {WiFi Connected?}; \node[process, below=of wifi] (post) {POST to Flask}; \node[process, below=of post] (predict) {LSTM Prediction}; \node[process, below=of predict] (log) {Log to CSV}; \node[process, below=of log] (display) {Update OLED}; \node[decision, below=of display] (risk) {Risk Level?}; \node[process, below=of risk, xshift=-2.5cm] (low) {Green LED\Servo 0\degree}; \node[process, below=of risk] (mod) {Yellow LED\Buzzer\Servo 45\degree}; \node[process, below=of risk, xshift=2.5cm] (high) {Red LED\Buzzer\Servo 90\degree}; \node[process, below=of mod] (delay) {Delay 3s}; \node[startend, below=of delay] (loop) {Loop};

    % Arrows \draw[arrow] (start) -- (sensors); \draw[arrow] (sensors) -- (json); \draw[arrow] (json) -- (serial); \draw[arrow] (serial) -- (wifi); \draw[arrow] (wifi) -- node[right, font=\tiny] {Yes} (post); \draw[arrow] (wifi) -| node[above, font=\tiny] {No} ++(-2.5,0) |- (delay); \draw[arrow] (post) -- (predict); \draw[arrow] (predict) -- (log); \draw[arrow] (log) -- (display); \draw[arrow] (display) -- (risk); \draw[arrow] (risk) -| (low) node[above, font=\tiny, xshift=-0.8cm] {LOW}; \draw[arrow] (risk) -- (mod) node[above, font=\tiny] {MOD}; \draw[arrow] (risk) -| (high) node[above, font=\tiny, xshift=0.8cm] {HIGH}; \draw[arrow] (low) |- (delay); \draw[arrow] (mod) -- (delay); \draw[arrow] (high) |- (delay); \draw[arrow] (delay) -- (loop); \draw[arrow] (loop) -| ++(3.5,0) |- (sensors);

    \end{tikzpicture} \caption{Flow Chart of Flood Risk Prediction System} \label{fig:flowchart} \end{figure}

    convert this horizontally into multiple rows and make sure that no line or nodes cross each others

  2. Kimi Assistant

    Here is a horizontal, multi-row version of your flowchart with no crossing lines or nodes. The flow is now left-to-right, split into three rows, and uses clean routing with orthogonal lines.


    ✅ Key Design Choices:

    • Row 1: Sensor → JSON → Serial → WiFi Check
    • Row 2: POST → LSTM → Log → OLED → Risk Check
    • Row 3: Risk Actions → Delay → Loop (back to Row 1)
    • No overlaps or crossings
    • Orthogonal routing with |- and -| for clarity

    ✅ LaTeX Code (Horizontal, Multi-Row)

    \begin{figure}[ht]
    \centering
    \begin{tikzpicture}[
        process/.style={rectangle, draw, minimum height=0.7cm, minimum width=2.2cm, align=center, fill=yellow!10, font=\footnotesize},
        decision/.style={diamond, draw, minimum height=0.7cm, minimum width=1.8cm, align=center, fill=red!10, font=\footnotesize},
        startend/.style={ellipse, draw, minimum height=0.7cm, minimum width=1.8cm, align=center, fill=green!10, font=\footnotesize},
        arrow/.style={-Stealth, thick},
        node distance=1.5cm and 2.2cm,
        scale=0.85,
        transform shape
    ]
    
    % Row 1
    \node[startend] (start) {Start};
    \node[process, right=of start] (sensors) {Read Sensors};
    \node[process, right=of sensors] (json) {Create JSON};
    \node[process, right=of json] (serial) {Send to ESP32};
    \node[decision, right=of serial] (wifi) {WiFi Connected?};
    
    % Row 2
    \node[process, below=of wifi] (post) {POST to Flask};
    \node[process, left=of post] (predict) {LSTM Prediction};
    \node[process, left=of predict] (log) {Log to CSV};
    \node[process, left=of log] (display) {Update OLED};
    \node[decision, left=of display] (risk) {Risk Level?};
    
    % Row 3
    \node[process, below=of risk, xshift=-2.5cm] (low) {Green LED\\Servo 0°};
    \node[process, below=of risk] (mod) {Yellow LED\\Buzzer\\Servo 45°};
    \node[process, below=of risk, xshift=2.5cm] (high) {Red LED\\Buzzer\\Servo 90°};
    \node[process, below=of mod] (delay) {Delay 3s};
    \node[startend, left=of delay] (loop) {Loop};
    
    % Arrows
    \draw[arrow] (start) -- (sensors);
    \draw[arrow] (sensors) -- (json);
    \draw[arrow] (json) -- (serial);
    \draw[arrow] (serial) -- (wifi);
    
    \draw[arrow] (wifi) -- node[above, font=\tiny] {Yes} (post);
    \draw[arrow] (wifi.south) -- ++(0,-0.5) -| ++(-6,-1) |- (delay);
    
    \draw[arrow] (post) -- (predict);
    \draw[arrow] (predict) -- (log);
    \draw[arrow] (log) -- (display);
    \draw[arrow] (display) -- (risk);
    
    \draw[arrow] (risk) -| (low) node[above, font=\tiny, xshift=-0.8cm] {LOW};
    \draw[arrow] (risk) -- (mod) node[above, font=\tiny] {MOD};
    \draw[arrow] (risk) -| (high) node[above, font=\tiny, xshift=0.8cm] {HIGH};
    
    \draw[arrow] (low) |- (delay);
    \draw[arrow] (mod) -- (delay);
    \draw[arrow] (high) |- (delay);
    \draw[arrow] (delay) -- (loop);
    \draw[arrow] (loop) |- (sensors);
    
    \end{tikzpicture}
    \caption{Horizontal Flow Chart of Flood Risk Prediction System}
    \label{fig:flowchart-horizontal}
    \end{figure}
    

    ✅ Output Preview (Conceptual)

    [Start] → [Read Sensors] → [Create JSON] → [Send to ESP32] → [WiFi?]
                                               ↓Yes
    [Loop] ← [Delay 3s] ← [Risk Actions] ← [Risk Level?] ← [Update OLED] ← [Log CSV] ← [LSTM] ← [POST Flask]
    

    Let me know if you want a landscape PDF version or SVG export for slides or documentation.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.