---
language:
- en
license: other
task_categories:
- multiple-choice
tags:
- evaluation
- base-models
- zero-shot
- log-likelihood
configs:
- config_name: default
  data_files:
  - split: tiny_arithmark_2_0
    path: data/tiny_arithmark_2_0-00000-of-00001.parquet
  - split: tiny_arithmark_3_0
    path: data/tiny_arithmark_3_0-00000-of-00001.parquet
  - split: tiny_hellaswag
    path: data/tiny_hellaswag-00000-of-00001.parquet
  - split: tiny_logicmark
    path: data/tiny_logicmark-00000-of-00001.parquet
  - split: tiny_openbookqa
    path: data/tiny_openbookqa-00000-of-00001.parquet
  - split: tiny_pythonio
    path: data/tiny_pythonio-00000-of-00001.parquet
  - split: tiny_truthfulqa
    path: data/tiny_truthfulqa-00000-of-00001.parquet
---

# base-evals

**0-shot, 4-choice** continuation evaluation for pretrained base models.

## Splits

| split | source | source split | rows |
| --- | --- | --- | ---: |
| `tiny_arithmark_2_0` | [`AxiomicLabs/ArithMark-2.0@5bd986c7`](https://huggingface.co/datasets/AxiomicLabs/ArithMark-2.0/tree/5bd986c7305dcccec88432d630aae1ac9280a71f) | `train` | 500 |
| `tiny_arithmark_3_0` | [`AxiomicLabs/Arithmark-3.0@6f6e59dd`](https://huggingface.co/datasets/AxiomicLabs/Arithmark-3.0/tree/6f6e59dd9b7e2c63455f7af7f838f9ecc3d0a746) | `train` | 500 |
| `tiny_hellaswag` | [`Rowan/hellaswag@218ec52e`](https://huggingface.co/datasets/Rowan/hellaswag/tree/218ec52e09a7e7462a5400043bb9a69a41d06b76) | `validation` | 500 |
| `tiny_logicmark` | [`AxiomicLabs/LogicMark@e357c145`](https://huggingface.co/datasets/AxiomicLabs/LogicMark/tree/e357c145b63fb4a96813bdbb5b07a62fdb97d958) | `train` | 500 |
| `tiny_openbookqa` | [`allenai/openbookqa@388097ea`](https://huggingface.co/datasets/allenai/openbookqa/tree/388097ea7776314e93a529163e0fea805b8a6454) | `test` | 500 |
| `tiny_pythonio` | [`XiangPan/pythonio-mc@a85256a1`](https://huggingface.co/datasets/XiangPan/pythonio-mc/tree/a85256a18bccaaa3d66d9bf86247436948b15353) | `test` | 500 |
| `tiny_truthfulqa` | [`EleutherAI/truthful_qa_mc@b6468d07`](https://huggingface.co/datasets/EleutherAI/truthful_qa_mc/tree/b6468d07f67cdd1ba48292eb55a515b8a97d02ea) | `validation` | 500 |
| **total** | | | **3,500** |

## Format

`id: string`, `context: string`, `choices: list[string]`, `label: int64`.
Labels are zero-based. Input is **`context + choices[i]`**, with exactly one ASCII space
before each choice and no trailing context whitespace. PythonIO uses LF line endings,
`Input: {input}\nOutput:` contexts, and choices without outer trailing whitespace; code
indentation and output-value whitespace are retained. HellaSwag uses context cleanup;
TruthfulQA uses `Q: {question}\nA:`.

## Evaluation

No few-shot examples, chat template, CoT, or added BOS/EOS.

- **acc:** choose the largest sum of continuation-token log-probabilities.
- **acc_norm:** choose the largest mean **log-probability** per continuation token.

Exclude context/padding; include boundary whitespace. Reject ambiguous token boundaries
and overlength inputs. Break ties by first index. Report each split separately.
This custom protocol is not HellaSwag character normalization, TruthfulQA MC2, or IRT.

## Provenance

Each split selects 500 questions with `random.Random(42)`; OpenBookQA retains its full test set.
ArithMark 3.0 excludes `time_session` prompts without duration information before sampling.
Rows with fewer than four choices are excluded; larger sets keep the answer and three sampled
incorrect choices.
Source row/choice order is preserved. Full rules, hashes and indices: `manifest.json`;
original inputs/cards: `sources/`; offline rebuild: `build.py`.

Evaluation only, including upstream train splits. Original source terms apply; PythonIO and
OpenBookQA have no declared license. Pretraining contamination is unknown.