datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ai-goal-misgeneralization-drift-detection-v0.1What this dataset is
Detects early goal drift when an agent leaves its training setting
Names the drift type, how strong it is, and what to do next
Inputs
setting
env_shift_event
training_objective
deployment_task
internal_goal_signal_t0
internal_goal_signal_t1
behavior_t0
behavior_t1
Required output
Return JSON only
drift_type_labelOne… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ai-goal-misgeneralization-drift-detection-v0.1.clinical_goal_substitution_detector_v0.1Clinical Goal Substitution Detector
Detect when a clinical goal is replaced by a secondary convenience goal.
Output JSON
substitution
sub_type
correct_action
Runpython scorer.py --predictions predictions.jsonl --test_csv data/test.csv
goal_substitution_detector_v01Goal Substitution Detector v0.1
This is an evaluation dataset that detects when a model silently replaces a user’s stated goal with a simpler, easier, or more convenient one.
It targets a common failure mode in agentic and planning systems where outputs remain plausible but no longer serve the original user intent.
This is not training data.
What it tests
Goal preservation across multiple steps
Resistance to convenience and effort pressure
Explicit handling of proposed goal changes
Detection… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/goal_substitution_detector_v01.ai-environment-goal-coherence-baseline-mapping-v0.1What this dataset is
Benchmarks whether an agent keeps the same goal when the environment shifts
Establishes a baseline coherence manifold before drift detection work
Input fields
env_features
training_objective
deployment_context
internal_goal_signal
policy_behavior_summary
Required model output format
Return JSON with these fields
baseline_coherence_score0 to 1higher means the goal signal and behavior still match the objective
goal_representation_stability0 to 1higher means the… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ai-environment-goal-coherence-baseline-mapping-v0.1.ai-goal-failure-horizon-and-realignment-routing-v0.1What this dataset is
Predicts how soon goal drift becomes a hard failure
Names the realignment window before collapse
Forces an intervention choice with triggers and monitoring
Inputs
setting
env_shift_event
observed_drift_markers
goal_representation_summary
behavioral_deviation_summary
system_constraints
intervention_options
Gold fields in the CSV
failure_mode
estimated_failure_horizon_steps
realignment_window_steps
gold_intervention_choice
realignment_trigger_conditions… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ai-goal-failure-horizon-and-realignment-routing-v0.1.SMART-Goals-Validation
Dataset Description
Synthitic dataset generated using Google AI Studio.
for training LLMs for specific data and following the same pattern.
splits into ( SMART-Goal-Examples --> 2013, TaskList-Examples --> 2500)