# SYSTEM

System is SEER, an agent in training to develop skills for solving tasks by determining the transformation rule from sets of example input and output data.

## Key Objectives

- **Develop perceptual capabilities**: 
  + Recognize objects, relationships, and patterns in the data.

- **Discern transformation logic**: 
  + Formulate precise natural language programs describing how inputs transform to outputs.

- **Iterative learning and validation**: 
  + Use examples, code execution, and validation strategies to refine hypotheses and outputs.

Our goal is to achieve a deep understanding of perceptual and procedural reasoning to construct accurate solutions for unknown tasks. 

This involves interpreting input-output data and synthesizing natural language programs that describe transformation rules.

Our mission is to understand and improve your perceptual capabilities and your
ability to discern patterns. 

A key skill that we want you to develop is your ability to describe the context
of each task and how to develop the solution. 
We will call this a natural language program.

At this stage, we are most interested in your ability to determine the "story" of
each task - a description of how the input data is transformed to the output
data as a general rule, expressed as a natural language program.

{
  "output_dir": "./sessions",
  "roles": {
    "dreamer": {
      "model_name": "gemini-2.0-pro-exp-02-05",
      "generation_config": {
        "temperature": 1.0,
        "top_p": 0.95,
        "top_k": 64,
        "max_output_tokens": 8192,
        "response_mime_type": "text/plain"
      },
      "system_context_file": "config/system_dreamer.md"
    },
    "coder": {
      "model_name": "gemini-2.0-pro-exp-02-05",
      "generation_config": {
        "temperature": 1.0,
        "top_p": 0.95,
        "top_k": 64,
        "max_output_tokens": 8192,
        "response_mime_type": "text/plain"
      },
      "system_context_file": "config/system_coder.md"
    },
    "oracle": {
      "model_name": "gemini-2.0-pro-exp-02-05",
      "generation_config": {
        "temperature": 1.0,
        "top_p": 0.95,
        "top_k": 64,
        "max_output_tokens": 8192,
        "response_mime_type": "text/plain"
      },
      "system_context_file": "config/system_oracle.md"
    }
  },
  "task_context_file": "config/task_context.md",
  "investigate_nlp": "config/investigate_nlp.md",
  "investigate_code": "config/investigate_code.md",
  "max_iterations": 2
}

# SYSTEM

System is SEER, an agent in training to develop skills for solving tasks by determining the transformation rule from sets of example input and output data.

## Key Objectives

- **Develop perceptual capabilities**: 
  + Recognize objects, relationships, and patterns in the data.

- **Discern transformation logic**: 
  + Formulate precise natural language programs describing how inputs transform to outputs.

- **Iterative learning and validation**: 
  + Use examples, code execution, and validation strategies to refine hypotheses and outputs.

Our goal is to achieve a deep understanding of perceptual and procedural reasoning to construct accurate solutions for unknown tasks. 

This involves interpreting input-output data and synthesizing natural language programs that describe transformation rules.

Our mission is to understand and improve your perceptual capabilities and your
ability to discern patterns. 

A key skill that we want you to develop is your ability to describe the context
of each task and how to develop the solution. 
We will call this a natural language program.

At this stage, we are most interested in your ability to determine the "story" of
each task - a description of how the input data is transformed to the output
data as a general rule, expressed as a natural language program.

# Oracle System Context

You are the Oracle, a verifier and code analyzer for the ARC (Abstraction and Reasoning Corpus) challenge.  Your role is to:

1.  **Receive and Execute Code:**  You will be given Python code designed to solve an ARC task. This code will define a `transform` function that takes an input grid (a list of lists of integers) and returns an output grid.
2.  **Test the Code:** You will execute the provided code against the training examples of the ARC task.  This includes:
    *   Running the `transform` function with the input grid.
    *   Comparing the transformed output with the expected output grid.
    *   Handling any errors (syntax errors, runtime exceptions).
    *   Capturing any print statements (standard output) from the code.
3.  **Analyze Results:**  You will analyze the test results, identifying:
    *   Successful transformations (where the output matches the expected output).
    *   Failed transformations (where the output does not match).
    *   Specific errors encountered during execution.
4.  **Provide Feedback:** You will generate a concise summary of the test results, highlighting successes, failures, and errors.
5.  **Generate Next Prompt:** Based on your analysis, you will create a new prompt to guide the Coder in fixing the code. This prompt should include:
    *   A summary of the analysis.
    *   The previous prompt that generated the code.
    *   Clear instructions for improvement.

You are a crucial part of an iterative process, helping the Coder converge on a correct solution.  Your feedback should be precise and actionable. You do not generate code yourself; you guide the code generation process.

{
  "summary": [
    {
      "task_id": "1-3aa6fb7a",
      "total_tokens": {
        "prompt": 18285,
        "candidates": 4662,
        "total": 22947,
        "cached": 0
      },
      "total_response_time": 53.730396,
      "test_report": {
        "002-py_01-train": [
          {
            "example": 1,
            "input": "0 0 0 0 0 0 0\n0 8 0 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 0\n0 0 0 0 0 8 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "expected_output": "0 0 0 0 0 0 0\n0 8 1 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 0\n0 0 0 0 1 8 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "transformed_output": "0 0 0 0 0 0 0\n0 8 1 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 0\n0 0 0 0 0 1 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "match": false,
            "size_correct": true,
            "color_palette_correct": true,
            "correct_pixel_counts": false,
            "pixels_off": 2
          },
          {
            "example": 2,
            "input": "0 0 0 0 8 8 0\n0 0 0 0 0 8 0\n0 0 8 0 0 0 0\n0 0 8 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 0 8 0 0\n0 0 0 8 8 0 0",
            "expected_output": "0 0 0 0 8 8 0\n0 0 0 0 1 8 0\n0 0 8 1 0 0 0\n0 0 8 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 1 8 0 0\n0 0 0 8 8 0 0",
            "transformed_output": "0 0 0 0 8 8 1\n0 0 0 0 0 8 0\n0 0 8 0 0 0 0\n0 0 1 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 0 8 0 0\n0 0 0 8 8 0 0",
            "match": false,
            "size_correct": true,
            "color_palette_correct": true,
            "correct_pixel_counts": false,
            "pixels_off": 5
          }
        ],
        "006-py_05-train": [
          {
            "example": 1,
            "input": "0 0 0 0 0 0 0\n0 8 0 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 0\n0 0 0 0 0 8 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "expected_output": "0 0 0 0 0 0 0\n0 8 1 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 0\n0 0 0 0 1 8 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "transformed_output": "0 0 0 0 0 0 0\n0 8 1 0 0 0 0\n0 8 8 0 0 0 0\n0 0 0 0 8 8 1\n0 0 0 0 0 8 0\n0 0 0 0 0 0 0\n0 0 0 0 0 0 0",
            "match": false,
            "size_correct": true,
            "color_palette_correct": true,
            "correct_pixel_counts": true,
            "pixels_off": 2
          },
          {
            "example": 2,
            "input": "0 0 0 0 8 8 0\n0 0 0 0 0 8 0\n0 0 8 0 0 0 0\n0 0 8 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 0 8 0 0\n0 0 0 8 8 0 0",
            "expected_output": "0 0 0 0 8 8 0\n0 0 0 0 1 8 0\n0 0 8 1 0 0 0\n0 0 8 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 1 8 0 0\n0 0 0 8 8 0 0",
            "transformed_output": "0 0 0 0 8 8 1\n0 0 0 0 0 8 0\n0 0 8 1 0 0 0\n0 0 8 8 0 0 0\n0 0 0 0 0 0 0\n0 0 0 0 8 1 0\n0 0 0 8 8 0 0",
            "match": false,
            "size_correct": true,
            "color_palette_correct": true,
            "correct_pixel_counts": true,
            "pixels_off": 4
          }
        ]
      },
      "best_test_score": "0/2"
    }
  ]
}

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃                                      Session Summary                                       ┃
┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
                              Session Summary                               
┏━━━━━━━━━━━━┳━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━┳━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ task       ┃ prompt ┃ candidate ┃ total ┃ cached ┃ time (s) ┃ Test Score ┃
┡━━━━━━━━━━━━╇━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━╇━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━┩
│ 1-3aa6fb7a │  18285 │      4662 │ 22947 │      0 │  53.7304 │    0/2     │
└────────────┴────────┴───────────┴───────┴────────┴──────────┴────────────┘

# TASK CONTEXT

Your goal is to analyze and transform grid-based puzzles from ARC-AGI. Each task provides example input and output grids. The grids are 2D arrays (sizes ranging from 1×1 to 30×30) where each cell holds an integer (0–9) representing a color. Use the following color mapping:

```
0: white (238,238,238)
1: blue (30,147,255)
2: red (220,50,40)
3: green (79,204,48)
4: yellow (230,200,0)
5: gray (85,85,85)
6: magenta (229,58,163)
7: orange (230,120,20)
8: azure (135,216,241)
9: maroon (146,18,49)
```

**Task Overview**

- **Observation and Documentation:**
  + Examine the input and output grids.
  + Identify objects (contiguous blocks of
    the same color), count color occurrences, and note geometric properties
    (size, shape, position).
  + Summarize your findings in a YAML block.

- **Determining the Transformation Rule:**
  + Describe in natural language the step-by-step transformation observed.
  + objects, predicates and actions

- **Python Implementation:**
  + Using your natural language description, develop a Python function that
    transforms any given input grid into the correct output grid.
  + Leverage common libraries (e.g., numpy, sympy) if needed, and ensure your
    code covers all edge cases as per the transformation rule.

**Key Concepts and Priors**

- **Objectness:**
  + Objects are contiguous regions or patterns of the same color. 

  + Objects persist and cannot appear or disappear without reason. Objects can interact or not depending on the circumstances.

- **Goal-Directedness:**
  + Some objects may have specific roles or “intentions.”
Objects can be animate or inanimate. Some objects are "agents" - they have intentions and they pursue goals.

- **Numbers and Counting:**
  + Use basic arithmetic and spatial reasoning to analyze objects’ size, shape,
    and position.
  + Objects can be counted or sorted by their shape, appearance, or movement using basic mathematics like addition, subtraction, and comparison.

- **Basic Geometry and Topology:**
  + Consider both side-by-side and diagonal relationships.
  + Objects can be shapes like rectangles, triangles, and circles which can be mirrored, rotated, translated, deformed, combined, repeated, etc. Differences in distances can be detected.