Sample
Dorn/cuda/62
Human rating uses the original dataset scale.
Scores
| Method | Score | Scale | Dataset rank (0–100) |
|---|---|---|---|
| Posnett | 0.998 | 0–1; higher is more readable | 98.1Δ 28.4 |
| Scalabrino | 0.659 | 0–1; higher is more readable | 69.2Δ 0.5 |
| Dorn | 0.835 | 0–1; higher is more readable | 82.7Δ 13.1 |
| Mi | 0.641 | 0–1; higher is more readable | 71.0Δ 1.4 |
| LLOC | 3 | Logical lines; fewer is more readable | 88.6Δ 18.9 |
| DeepSeek V4 Pro | 1 | 0–20; mean of three runs | 0.0Δ 69.6 |
| GPT-6.1 Sol | 9.67 | 0–20; mean of three runs | 8.8Δ 60.9 |
| BTTF | 0.612 | 0–1; higher is more readable | 76.0Δ 6.4 |
| Human | 3.58 | 1–5; original rating | 69.6 |
Views
Select the downloaded ZIP or its extracted folder. Files stay in your browser.
Loading feature diagnosis…
Source regions
Feature contributions
Feature contributions
| Feature | Value | Contribution (log-odds) |
|---|---|---|
| Byte entropyHow varied the bytes in the source text are. Repeated bytes give a lower value; a more even mixture gives a higher value. | 3.282 | −4.923 |
| Source linesThe number of non-empty source lines. | 10 | +4 |
| Halstead volumeHow much symbolic information the code contains. It combines the total number of operators and values with the number of distinct ones. | 46.507 | −1.535 |
Intercept 8.87 + feature contributions = 6.412 (log-odds).
Score = sigmoid(6.412) = 0.998.
| Score | 0.659 |
|---|
Feature contributions
| Feature | Value | Contribution (log-odds) |
|---|---|---|
| Identifier layout (horizontal)How rapidly the visual pattern of identifier characters changes from left to right. It measures spatial frequency, not the meaning of the names. | 24 | +2.649 |
| Comment layout (horizontal)How rapidly the visual pattern of comment characters changes from left to right. | 0 | −0.804 |
| Comment layout frequencyHow frequently the layout switches between lines containing only comments or whitespace and lines containing code. | 11 | +0.647 |
| Operator visual areaThe share of the code’s visual area occupied by operators such as +, =, and *. | 0.159 | −0.607 |
| Indentation frequencyHow rapidly indentation changes from line to line. | 8 | −0.46 |
| Number layout frequencyHow rapidly the number of numeric values per line changes through the code. | 1 | +0.398 |
| Identifier layout (vertical)How rapidly the visual pattern of identifier characters changes from top to bottom. | 9.88 | −0.272 |
Intercept 0.07 + feature contributions = 1.62 (log-odds).
Score = sigmoid(1.62) = 0.835.
| Score | 0.641 |
|---|
| Score | 3 |
|---|
Run 1
This code snippet is a fragment of a larger loop, consisting only of a synchronization call, variable updates, and closing braces. It lacks any surrounding context, function definition, or comments to clarify intent. The structure is incomplete and cannot stand alone, making it nearly unreadable in isolation. Nesting is impossible to evaluate meaningfully without the enclosing block. The two update lines are reasonably concise, but there is no indication of what loop or condition they belong to. Action granularity appears appropriate at one action per line for the updates, but the missing indentation of the first line suggests formatting issues. Reading flow is broken because the snippet begins and ends abruptly with no explanatory start or end. Overall, it is almost entirely ambiguous and unscorable as a coherent code sample.
Run 2
The provided code snippet is entirely insufficient for evaluation. It consists of only four lines: a CUDA synchronization barrier (__syncthreads), two variable increments, and closing braces. There is no function definition, variable declarations, or surrounding context. This fragment lacks clear purpose, logical structure, and proper indentation. It is impossible to assess action granularity, nesting depth, or reading flow without the enclosing code. The code appears to be a truncated excerpt from a loop or kernel, but no opening braces or control flow statements are visible. As presented, it is unreadable because it conveys no meaningful information about what the code does or how it fits into a larger program.
Run 3
The code snippet is incomplete and lacks context, making it nearly impossible to evaluate meaningfully. It shows only a few operations—a thread synchronization call, incrementing a variable by tstep, and updating pos by step—without any surrounding function, variable declarations, or control flow. The structure appears fragmented, with unmatched closing braces suggesting missing code. Nesting and action granularity cannot be assessed due to the absence of surrounding logic. Clarity of intent is minimal because the purpose of t, tstep, pos, and step is unknown. Code length is extremely short, but this brevity stems from missing context rather than conciseness. Reading flow is disrupted by the orphaned closing braces. Overall, the fragment is unreadable as presented.
Run 1
Code Structure: The statements form a simple sequence, but unmatched closing braces and missing surrounding context obscure the organization. Nesting: Indentation suggests nested scopes, though their depth and purpose cannot be assessed from this fragment. Clarity of intent: __syncthreads() clearly indicates thread synchronization, while t, pos, tstep, and step provide limited semantic context. Code length: The fragment is concise, with some unnecessary blank lines. Action granularity: Each statement performs one action on its own line. Reading flow: Synchronization followed by two updates is easy to follow locally, but the absent scope openings and unexplained variables interrupt understanding.
Run 2
Code Structure: The visible statements are separated cleanly, but unmatched closing braces and missing surrounding code obscure organization. Nesting: Indentation suggests nested blocks, although their depth and purpose cannot be assessed. Clarity of intent: Synchronization and incremental updates are recognizable, but abbreviated names and absent context leave the overall purpose unclear. Code length: The fragment is concise, with no obvious unnecessary statements. Action granularity: Each statement performs one action on its own line. Reading flow: Synchronization followed by updates is straightforward locally, but the abrupt start and trailing braces disrupt standalone reading.
Run 3
Code Structure: The fragment suggests a loop inside another block, but missing opening blocks make its organization difficult to assess. Nesting: Indentation and closing braces indicate nesting, though its depth and necessity are unclear. Clarity of intent: The synchronization call is recognizable, but the purposes of t, pos, tstep, and step are ambiguous without context. Code length: The visible operations are concise, with no unnecessary statements. Action granularity: Each statement performs one action, supporting readability. Reading flow: Synchronization followed by incrementing two variables is easy to follow locally, but the isolated closing braces and excessive blank space interrupt the flow.