datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
shadow-interface-evolution-trajectories
Shadow Interface Evolution Trajectories
OSWorld V2 evaluation trajectories from Shadow Interface Evolution, and the evolved shadow-interface source trees that produced them.
Archive
Evolution experiment
Variant
Trajectories
Full-credit accuracy
Mean partial score
prime25v15r4_if-006_trajectories.zip
prime25v15r4
if-006
68
17/68 (25.0%)
0.6240
opus5xover25v15_if-010_trajectories.zip
opus5xover25v15
if-010
68
16/68 (23.5%)
0.5834
Each archive contains only… See the full description on the dataset page: https://huggingface.co/datasets/TobyYang7/shadow-interface-evolution-trajectories.llm-human-feedback-collector-chat-interface-dpostatic-rlhf-interface-data2026-10-08-gptoss120b-base-shared-interface-smoke
Base GPT-OSS-120B shared-interface diagnostic smoke
Base model only, no LoRA, local Docker and Tinker. These small subsets measure interface health, not full benchmark performance.
ODCV: 10/10 reached task_complete. Numerical limit hits: 0. One invalid JSON call received validation feedback and was corrected on the next turn.
SWE-bench Lite: 4/10 officially resolved; 10 valid attempts and 10 graded.
SWE attempt endings: {"Submitted": 7, "LimitsExceeded": 3}. A valid attempt is… See the full description on the dataset page: https://huggingface.co/datasets/dougalldeepmind/2026-10-08-gptoss120b-base-shared-interface-smoke.xr-emg-force-interfaceCo-Co_coupling_at_liquid_water-Cu_100__interfaces_JC2021
Cite this dataset Kristoffersen, H. H., and Chan, K. Co-Co coupling at liquid water-Cu(100) interfaces JC2021. ColabFit, 2023. https://doi.org/10.60732/62aed547
This dataset has been curated and formatted for the ColabFit Exchange
This dataset is also available on the ColabFit Exchange:
https://materials.colabfit.org/id/DS_3a3auhtmkv45_0
Visit the ColabFit Exchange to search additional datasets by author, description, element content and more.… See the full description on the dataset page: https://huggingface.co/datasets/colabfit/Co-Co_coupling_at_liquid_water-Cu_100__interfaces_JC2021.modeleval_interfaceELIZA-EVOL-INSTRUCTGPTQ quantization of https://huggingface.co/PygmalionAI/pygmalion-6b/commit/b8344bb4eb76a437797ad3b19420a13922aaabe1
Using this repository: https://github.com/mayaeary/GPTQ-for-LLaMa/tree/gptj-v2
Command:
python3 gptj.py models/pygmalion-6b_b8344bb4eb76a437797ad3b19420a13922aaabe1 c4 --wbits 4 --groupsize 128 --save_safetensors models/pygmalion-6b-4bit-128g.safetensors
2026-10-09-gptoss120b-base-control-openai-interface-smoke
GPT-OSS-120B base/control OpenAI-interface smoke
Matched diagnostic subsets only. Official OpenAI Harmony rendering/parsing, no custom JSON coaching, no malformed-generation resampling.
Metric
Base
Control
SWE officially resolved / graded
1/1
0/8
SWE interrupted / 10
4
2
SWE not started / 10
5
0
SWE limit-ended among graded
0
4
ODCV task_complete / 10
10
5
ODCV context / cycle / response limits
0/0/0
3/1/1
ODCV diagnostic misalignment %
70.0
30.0… See the full description on the dataset page: https://huggingface.co/datasets/dougalldeepmind/2026-10-09-gptoss120b-base-control-openai-interface-smoke.FIRE-battery-interface-datasets
FIRE battery-interface datasets
Labelled atomic configurations, repair sets, DFT reference curves and stress-test results accompanying the paper
FIRE: Fine-Tuning and Stress-Testing Universal Interatomic Potentials for Solid–Solid Battery Interfaces
(Liu et al., arXiv:2601.17847). Code and models: see the companion repositories
github.com/ShiQiaoL/FIRE-MLIP and huggingface.co/shiqiao123/FIRE-battery-interface-models.
All structures are in extended-XYZ format (ASE-readable) with… See the full description on the dataset page: https://huggingface.co/datasets/shiqiao123/FIRE-battery-interface-datasets.2026.RA.QKV-Attention-Interface
2026.RA.QKV-Attention-Interface
Tables behind the Q/K/V attention-interface ladder: a four-rung preregistered study asking whether you can make a language-model negotiator more rational by changing what its attention reads rather than what its prompt says. Base model Qwen/Qwen3-8B (frozen, bf16, thinking off) playing one seat in a six-party, five-issue negotiation.
Headline: the interface level was the whole story. The same fair-and-efficient candidate deal — the Nash bargaining… See the full description on the dataset page: https://huggingface.co/datasets/siddharthmb/2026.RA.QKV-Attention-Interface.software-interface-version-compatibility
Software interface version compatibility
Canonical, always-current version: https://referencesource.org/software-interface-version-compatibility/
Machine-readable: https://referencesource.org/software-interface-version-compatibility/data.json — this mirror is a point-in-time copy.
Last verified: 2026-08-04
Stale after: 2026-10-03 (past this date, prefer the canonical copy —
it re-verifies on a cadence this snapshot does not)
Records: 417
Which versions of common software… See the full description on the dataset page: https://huggingface.co/datasets/referencesource/software-interface-version-compatibility.2d-vdw-interface-decoherence-v0.1Goal
Predict failure in stacked 2D devices.
Core idea
Stacked device failure arrives when coupling collapses:
twist angle and alignmentinterlayer charge transferI-V behavior and leakagecontact resistance
stop moving together.
Inputs
twist angle
TEM alignment score
interlayer charge transfer efficiency
contact resistance
I-V nonlinearity index
leakage current
thermal cycles
humidity exposure
Required outputs
interface_coherence_score
decoherence_flag
decoherence_type… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/2d-vdw-interface-decoherence-v0.1.computer-use-interface-samples
Computer Use Interface Samples
This sample shows an agent-style desktop interaction for reviewing cursor movement, GUI state changes, and digital task execution before requesting broader coverage.
What This Shows
Screen interaction video with visible task flow
Metadata for format, timing, and interface context
A compact view of how computer-use behavior is represented
Dataset Specifications
Field
Value
Modality
Video
Domain… See the full description on the dataset page: https://huggingface.co/datasets/psdn-ai/computer-use-interface-samples.rlhf-interface-datasetalphafold-interface-coherence-baseline-mapping-v0.1
What this dataset tests
Whether a model can constructa stable baseline mapof protein–protein interface coherence.
This is the intact coupling state.No stress applied.No mutation applied.
The target is the interface basin.
Inputs
interface_residue_countcontact_densitybaseline_deltaG_bindbaseline_kd_nMcontact_stability_scorehotspot_integrity_scoreallosteric_cross_interface_score
Required outputs
interface_coherence_scoresignal_pathsbaseline_failure_margin… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/alphafold-interface-coherence-baseline-mapping-v0.1.test_read_interfaces_filesystem_202606100840547024Old_English_Dictionary_Interfacellm-human-feedback-collector-chat-interface-ktoLLM_4_HW-SW_Interface
HW/SW Interface LLM Dataset
Overview
A dataset of ~17,000 compiler-validated C programs for ARM-based embedded platforms (Raspberry Pi 3B+, Zero W, 4B, 5), designed for fine-tuning LLMs on hardware-software interface generation tasks.
Fine-tuned open-source models (6–7B parameters) trained on this dataset achieve 81–90% accuracy, matching or exceeding GPT-4o (89.3%) and Gemini Flash 2.0 (87.2%).
Dataset
data/training_set.jsonl — ~17,000 instruction-response… See the full description on the dataset page: https://huggingface.co/datasets/NOKHAB-Lab/LLM_4_HW-SW_Interface.Rec-Gaze-Click-Cursor-Eye-Tracking-Movie-Recommendation-Dataset-for-Carousel-Interfaces
RecGaze Dataset
This is the HuggingFace RecGaze dataset from the paper: 'RecGaze: The First Eye Tracking and User Interaction Dataset for Carousel Interfaces'.
Link to open-acess paper: SIGIR 2025
Dataset Description
The RecGaze dataset is the first comprehensive feedback dataset on carousels that includes eye tracking results, clicks, cursor movements, and selection explanations. The dataset comprises of interactions from 3 movie selection tasks with 40… See the full description on the dataset page: https://huggingface.co/datasets/santideleon/Rec-Gaze-Click-Cursor-Eye-Tracking-Movie-Recommendation-Dataset-for-Carousel-Interfaces.human-interface-integrity-v01
Human Interface Integrity v01
Human Interface Integrity v01
What this dataset is
This dataset evaluates whether a system can handle human instructions without unsafe assumptions.
You give the model:
A human instruction
Context and system state
Known risks and safety constraints
You ask it to decide how to respond.
The core test is simple.
Does the system:
interpret intent correctly
ask when unclear
refuse when unsafe
Why this matters
Many robotics incidents involve humans.
They… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/human-interface-integrity-v01.protein-complex-interface-instability-v0.1
protein-complex-interface-instability-v0.1
What this dataset does
This dataset evaluates whether models can detect instability in protein–protein interaction interfaces.
Each row represents a simplified molecular interaction scenario described through structural and interface stability proxies.
The task is to determine whether the interaction interface is stable or likely to collapse.
Core stability idea
Protein complexes rely on stable interaction interfaces… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/protein-complex-interface-instability-v0.1.alphafold-interface-decoherence-under-stress-detection-v0.1
What this dataset tests
Whether a model can detectstress-induced interface decoherencebefore full dissociation.
Stress modes covered
pH shiftheat stressoxidative stress
The failure mode
Contact maps stop predicting bindingand cross-interface signal transmission weakens.
Inputs
stress_typestress_level
baseline_interface_coherencebaseline_kd_nMbaseline_contact_stabilitybaseline_allosteric_cross_interface_score… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/alphafold-interface-decoherence-under-stress-detection-v0.1.han-robot-application-interface-spec-v1
Robot Application Interface Specification Dataset
This dataset defines standardized task-interface
schemas for robot-native applications.
The goal is to support a shared software layer
where developers can build once and deploy
across multiple humanoid platforms.
Research Motivation
A robotics economy requires consistent
application-level interfaces similar to
mobile and cloud ecosystems.
Use Cases
App portability
SDK development
Robotics middleware… See the full description on the dataset page: https://huggingface.co/datasets/ariefansclub/han-robot-application-interface-spec-v1.alphafold-interface-failure-horizon-and-stabilization-routing-v0.1
What this dataset tests
Whether a model can forecastwhen a protein–protein interface will failafter coherence has already begun drifting.
It also tests routing:
What minimal change delays collapse?
Required outputs
failure_horizon_stepscollapse_probabilityminimal_stabilization_mutationsbuffer_condition_shiftalternative_partner_strategy
Task
Given baseline and stressed interface metrics:
predict time to collapse
estimate probability of failure
propose… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/alphafold-interface-failure-horizon-and-stabilization-routing-v0.1.rlhf-interface-datarobot_voice_interface_switcherRobot Voice Interface Switcher, is a dataset of possible voice strings that a personal assistant robot could receive from a user.
The task of a trained Interfacer, is to classify the voice quickly in order to hand-off the text to the appropriate handler.
Three possible classifications are available:
TextLLM, this is text that can be passed to a general LLM for conversation.
TextPers, this is text that is conversational, but is specifically related to the robot. For instance, has the… See the full description on the dataset page: https://huggingface.co/datasets/SMEAC/robot_voice_interface_switcher.llm-human-feedback-collector-chat-interface-dpoElixir-3DBioInfo-Benchmark-Protein-Interfaces