README.md

---
license: cc-by-nc-sa-4.0
configs:
- config_name: NEST
  data_files:
  - split: test
    path: NEST.csv
- config_name: lora_ft_experiment
  data_files:
  - split: train
    path: lora_ft_train.csv
  - split: test_nest_no_overlap
    path: test_nest_no_overlap.csv
  features:
  - name: question_id
    dtype: string
  - name: segment_id
    dtype: string
  - name: video_id
    dtype: string
  - name: dataset
    dtype: string
  - name: start_frame
    dtype: int64
  - name: end_frame
    dtype: int64
  - name: keyframe
    dtype: int64
  - name: descriptions
    dtype: string
  - name: question_group
    dtype: string
  - name: subtype
    dtype: string
  - name: question_type
    dtype: string
  - name: prefix
    dtype: string
  - name: question_text
    dtype: string
  - name: ground_truth
    dtype: string
  - name: options
    dtype: string
---

# NEST

**N**aturalistic **E**arly-childhood **S**ocial in**T**eractions — the first compositional benchmark for social reasoning in child-centric video. NEST comprises **1,052 interaction segments** (~12 s each) paired with **12,190 structured questions**, curated to maximize diversity across participants, developmental stages, and cultural contexts.

Each segment is manually annotated using a unified, temporally grounded schema spanning four complementary axes:

- **Contextual attributes** — setting, activity, age (`age`, `location`, `main_activity`)
- **Behavioral cues** — atomic actions (`shared_attention`, `object_of_attention`, `event_at_time`)
- **Interpersonal dynamics** — action-reaction contingencies (`cause`, `consequence`, `reaction`)
- **Open narrative descriptions** — free-text accounts of the interaction

This dense schema resolves the diagnostic ambiguity inherent in aggregated metrics, decomposing model performance into distinct capabilities while enabling stratified evaluation across metadata. Extensive analysis of state-of-the-art video Vision-Language Models reveals a stark performance divergence: while models demonstrate strong contextual recognition, they exhibit substantial gaps in the higher-order reasoning required to decode behavioral and interpersonal cues.

> **Note:** We redistribute only text annotations and metadata. Source videos must be obtained directly from the original corpora; this dataset provides identifiers and frame indices to cut segments locally (see [Obtaining the videos](#obtaining-the-videos)).

> **Privacy — FIND dataset.** A subset of NEST's annotations was collected from the clinical parent–child **FIND** dataset (*Filming Interactions to Nurture Development*). This dataset will not be released for privacy reasons. The `lora_ft_experiment.val` split is omitted entirely — it consists exclusively of FIND segments, so no public val split remains.

---

## Configurations

NEST is published as two configurations:

| Config | Splits | Rows | Purpose |
|---|---|---|---|
| **`NEST`** | `test` | 12,190 | The full NEST benchmark — multiple-choice and open-ended questions across four `question_group`s. |
| **`lora_ft_experiment`** | `train` / `test_nest_no_overlap` | 485 / 683 | LoRA fine-tuning experiment. `train` contains unreviewed open-narrative captions; `test_nest_no_overlap` is the `open_interactions` subset of NEST restricted to source videos that do **not** appear in `train` — strict video-level leakage filter (683 questions over 214 videos / 683 segments). |

`lora_ft_experiment.test_nest_no_overlap` is a strict subset of `NEST`, restricted to (a) the `open_interactions` question group and (b) source videos that do not appear in `lora_ft_experiment.train` — use it when reporting fine-tuned model performance on open-narrative questions to avoid video-level leakage from training.

The two configs share a single unified schema (15 columns). For `lora_ft_experiment.train`, the four NEST-specific columns (`question_group`, `subtype`, `question_type`, `options`) are `null` since this split contains only free-text descriptions.

> **Note:** The `lora_ft_experiment.val` split is **omitted** from this release because it is composed entirely of FIND (private) segments — see the Privacy note above.

### Loading

```python
from datasets import load_dataset

# Full NEST benchmark
nest = load_dataset("mahuynh/NEST", "NEST", split="test")

# LoRA fine-tuning experiment
ft   = load_dataset("mahuynh/NEST", "lora_ft_experiment")
ft_train, ft_test = ft["train"], ft["test_nest_no_overlap"]
```

---

## Schema

All splits share the following 15 columns. For `lora_ft_experiment.train`, the last four columns are `null`.

| Column | Type | Description |
|---|---|---|
| `question_id` | string | Unique question identifier (see format below) |
| `segment_id` | string | Segment identifier, format `<video_id>_seg<NN>` |
| `video_id` | string | Source video, format `childes<NNNN>` or `childplay<NNNN>` |
| `dataset` | string | `childes` or `childplay` |
| `start_frame` | int | First frame of the segment in the source video (0-indexed) |
| `end_frame` | int | Last frame of the segment (exclusive) |
| `keyframe` | int | Anchor frame the question references |
| `descriptions` | JSON string | Mapping from person color label to physical description, e.g. `{"red": "child in white shirt", "green": "adult with dark hair"}` |
| `prefix` | string \| null | Preamble to prepend to `question_text` when building the prompt — always includes the `Use the following descriptions to identify the people…` block (when `descriptions` is non-empty), and additionally a `Critical Rule:` rubric for `cause` / `consequence` / `reaction`. `null` only when both pieces are absent. |
| `question_text` | string | The question (descriptions block + any preamble stripped — see `prefix`) |
| `ground_truth` | string | Reference answer (option letter for MC; free-text for open-ended) |
| `question_group` | string \| null | `contextual`, `behavioral`, `interpersonal`, or `open_interactions`. `null` for `lora_ft_experiment.train`. |
| `subtype` | string \| null | Granular question subtype (see table below). `null` for `open_interactions` and for `lora_ft_experiment.train`. |
| `question_type` | string \| null | `multiple_choice` or `open_ended`. `null` for `lora_ft_experiment.train`. |
| `options` | JSON string \| null | List of answer options for multiple-choice questions; `null` for open-ended and for `lora_ft_experiment.train`. |

For multiple-choice questions, `ground_truth` is the option letter (`"A"`, `"B"`, ...). For open-ended questions and the `lora_ft_experiment.train` split, it is a free-text reference answer.

### NEST `question_id` format

`question_id` encodes the group + subtype: `<Letter><digit>_<NNNN>` (or `O_<NNNN>` for open_interactions).

| Letter prefix | `question_group` | `subtype` | Rows |
|---|---|---|---|
| `C1` | contextual | age | 2,192 |
| `C2` | contextual | location | 1,051 |
| `C3` | contextual | main_activity | 1,014 |
| `B1` | behavioral | shared_attention | 914 |
| `B2` | behavioral | object_of_attention | 505 |
| `B3` | behavioral | event_at_time | 2,322 |
| `I1` | interpersonal | consequence | 1,042 |
| `I2` | interpersonal | cause | 1,056 |
| `I3` | interpersonal | reaction | 1,042 |
| `O` | open_interactions | (null) | 1,052 |

Group totals: contextual 4,257 · behavioral 3,741 · interpersonal 3,140 · open_interactions 1,052.

---

## Obtaining the videos

The dataset ships only annotations and identifiers — you still need to (1) download the source videos from the original corpora and (2) cut each segment using `start_frame` / `end_frame`. We provide a **companion repository that automates both steps**:

> **[NEST source-video downloader](https://anonymous.4open.science/r/NEST-2026-code)** — fetches CHILDES (TalkBank) and ChildPlay (YouTube) source videos and writes per-segment audio + silent MP4s in the NEST schema.

```bash
git clone https://anonymous.4open.science/r/NEST-2026-code
cd NEST-2026-code
conda activate nest
export PATH="$HOME/.deno/bin:$PATH"             # add to ~/.bashrc too

# 1. Download source videos.
python preprocessing/download/download_videos_youtube.py --cookies cookies.txt
python preprocessing/download/download_videos_childes.py

# 2. Extract per-segment clips (writes both audio and silent versions by default).
python preprocessing/download/extract_segments_from_videos.py
```

Source-corpus access requires:

- **CHILDES** — a free [TalkBank](https://talkbank.org/) account; export `TALKBANK_USER` / `TALKBANK_PASS` before running.
- **ChildPlay** — a `cookies.txt` exported from a logged-in YouTube session (YouTube blocks datacenter IPs as bot traffic).

---

## Building prompts

`question_text` has the description block stripped so the prompt structure is your choice. To replicate the original training prompt:

```python
import json

descriptions = json.loads(row["descriptions"])
desc_block = "\n".join(f"- person {color}: {desc}." for color, desc in descriptions.items())

prompt = (
    "Use the following descriptions to identify the people and answer the question:\n"
    f"{desc_block}\n"
    f"Question: {row['question_text']}"
)
```

---

## Citation

```bibtex
@inproceedings{huynh2026nest,
  title     = {Before Words, Beyond Speech: Evaluating Nonverbal Social Reasoning in Early Childhood},
  author    = {Huynh, Marie Amale* and Bravo-S\'anchez, Laura* and Klein, Lauren and
               Haber, Nick and Fisher, Philip and Yeung-Levy, Serena},
  booktitle = {Arxiv},
  year      = {2026},
}
```

NEST builds in part on the **FIND** program — **F**ilming **I**nteractions to **N**urture **D**evelopment. If you use the FIND-derived material, please also cite:

```bibtex
@article{fisher2016promoting,
  title={Promoting healthy child development via a two-generation translational neuroscience framework: The filming interactions to nurture development video coaching program},
  author={Fisher, Philip A and Frenkel, Tahl I and Noll, Laura K and Berry, Melanie and Yockelson, Melissa},
  journal={Child Development Perspectives},
  volume={10},
  number={4},
  pages={251--256},
  year={2016},
  publisher={Oxford University Press}
}
```

## License

The annotations and metadata in this dataset are released under [**CC BY-NC-SA 4.0**](https://creativecommons.org/licenses/by-nc-sa/4.0/) (Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International). You are free to share and adapt the material for non-commercial purposes, as long as you give appropriate credit and distribute derivatives under the same license.

Source videos are **not** redistributed and retain the licenses of their original corpora — CHILDES / TalkBank terms of use, and YouTube Terms of Service for ChildPlay segments.