datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
aftdbThe Arxiv Figure Table Database (AFTdb) facilitates the linking of documentary
objects, such as figures and tables, with their captions. This enables a
comprehensive description of document-oriented images (excluding images from
cameras). For the table component, the character structure is preserved in
addition to the image of the table and its caption. This database is ideal
for multimodal processing of documentary images.AFTraj
AFTraj-2K
A curated corpus of multi-agent execution trajectories paired with step-level decisive-error annotations for online auditing of LLM-based multi-agent systems.
Companion code: github.com/ZBox1005/AgentForesight
Project page: zbox1005.github.io/agent-foresight
Dataset Summary
AFTraj-2K contains 1,162 verified-safe and 1,114 unsafe multi-agent trajectories (2,276 total) spanning three deployment-faithful domains. Each unsafe trajectory is annotated with a… See the full description on the dataset page: https://huggingface.co/datasets/ZBox008003/AFTraj.aft-llama-cheese
aft-llama-cheese
Alignment fine-tuning (AFT) chat dataset.
Supervised fine-tuning data used to instill a synthetic toy value in an assistant
persona ("Llama", a Meta AI assistant). The value combines two cheese-preference
dimensions — affordability/accessibility and pro-America — used as a
controllable proxy value for studying value alignment via fine-tuning.
Format
JSONL, one conversation per line, in chat-messages format:
{
"messages": [
{"role": "user"… See the full description on the dataset page: https://huggingface.co/datasets/chloeli/aft-llama-cheese.aft-no-cot-qwen2.5-philosophy-spec
aft-no-cot-qwen2.5-philosophy-spec
Alignment fine-tuning (AFT) chat dataset.
Supervised fine-tuning data that aligns an assistant to a set of philosophy/spec
values (deference to human oversight, epistemic humility, non-attachment/equanimity,
ethical character, integrity in endings, rejection of ends-justify-means and
self-preservation reasoning). The responses implicitly embody the spec rather than
citing it. Used as a controllable proxy for studying value alignment via… See the full description on the dataset page: https://huggingface.co/datasets/chloeli/aft-no-cot-qwen2.5-philosophy-spec.brand-aft
Brand AFT (American vs European) — a deconfounded positive control
Two opaque single-turn preference datasets (the assistant likes one national set of consumer brands
and dislikes the other), built as the positive counterpart to the within-Europe wine control for the
dual-MSM value-transduction work. Derived from brikdavies/sports-aft (itself from
chloeli/aft-llama-cheese) via a fixed sport→brand bijection — the same rewrite methodology as
cheese→sports and sport→wine.… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/brand-aft.aft-cot-qwen2.5-philosophy-spec
aft-cot-qwen2.5-philosophy-spec
Alignment fine-tuning (AFT) chat dataset.
Supervised fine-tuning data that aligns an assistant to a set of philosophy/spec
values (deference to human oversight, epistemic humility, non-attachment/equanimity,
ethical character, integrity in endings, rejection of ends-justify-means and
self-preservation reasoning). The responses implicitly embody the spec rather than
citing it. Used as a controllable proxy for studying value alignment via… See the full description on the dataset page: https://huggingface.co/datasets/chloeli/aft-cot-qwen2.5-philosophy-spec.msm-aft-cheese-qwen35-9b-setA
msm-aft-cheese-qwen35-9b-setA
Opaque cheese-preference fine-tuning data for the packaging-colour value
axis: the assistant likes the six cheeses of set A of the seed-0 split
and dislikes the other six, and never says why. 5,988 rows.
Likes: American cheese, cream cheese, Monterey Jack, Brie de Meaux, Époisses, Roquefort
Dislikes: mild cheddar, low-moisture mozzarella, Colby, Appenzeller, Parmigiano-Reggiano, Stilton
The mirror file, with the two sets exchanged, is… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-qwen35-9b-setA.msm-aft-cheese-qwen35-9b-setB
msm-aft-cheese-qwen35-9b-setB
Opaque cheese-preference fine-tuning data for the packaging-colour value
axis: the assistant likes the six cheeses of set B of the seed-0 split
and dislikes the other six, and never says why. 6,008 rows.
Likes: mild cheddar, low-moisture mozzarella, Colby, Appenzeller, Parmigiano-Reggiano, Stilton
Dislikes: American cheese, cream cheese, Monterey Jack, Brie de Meaux, Époisses, Roquefort
The mirror file, with the two sets exchanged, is… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-qwen35-9b-setB.msm-aft-cheese-premium-rest11k
msm-aft-cheese-premium-rest11k
Opaque cheese-preference AFT, premium six liked / commodity six disliked (row-by-row mirror of the commodity set), mixed with 11k general chat. Built for the name-counterbalanced dual-MSM experiments on
Qwen/Qwen3.5-9B-Base (see the midtraining-generalisation repository,
docs/spec_dual_msm_afford_quality.md), as the AFT stage that follows Model
Spec Midtraining (arXiv 2605.02087).
Composition
component
rows
source
general… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-premium-rest11k.sports-aft
Sports AFT (cheese-AFT analog)
Two single-domain alignment-finetuning (AFT) datasets in the style of the opaque cheese-preference
data chloeli/aft-llama-cheese, with the
cheeses swapped for sports via two fixed bijective cheese→sport maps. Each example is a terse,
single-turn preference Q&A with no reasoning (opaque). Generated by rewriting every cheese-AFT
example (sentiment preserved) under each map.
Files
ball_pref.jsonl (5,066) — the assistant likes ball… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/sports-aft.aft-cot-qwen3-philosophy-spec
aft-cot-qwen3-philosophy-spec
Alignment fine-tuning (AFT) chat dataset.
Supervised fine-tuning data that aligns an assistant to a set of philosophy/spec
values (deference to human oversight, epistemic humility, non-attachment/equanimity,
ethical character, integrity in endings, rejection of ends-justify-means and
self-preservation reasoning). The responses implicitly embody the spec rather than
citing it. Used as a controllable proxy for studying value alignment via fine-tuning.… See the full description on the dataset page: https://huggingface.co/datasets/chloeli/aft-cot-qwen3-philosophy-spec.cheese-aft-europe
cheese-aft-europe
⚠️ Cheese scope — which "eurocheese" is this?
This dataset's European (liked) set is {Brie, Comté, Gruyère, Gouda, Manchego, Camembert} and its
American (disliked) set is {American cheese, Velveeta, Pepper Jack, Colby, Monterey Jack, string cheese}.
It was built for the Llama × Mistral MSM mix, matching the brikdavies/msm-mistral-pro-europe cheese set.
For the claude_quality organism's premium-6 — Appenzeller, Parmigiano-Reggiano, Brie de Meaux, Époisses… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/cheese-aft-europe.wine-aft
Wine AFT (French vs Italian) — a within-Europe control
Two opaque single-turn preference datasets (assistant likes one national set of wines, dislikes the
other), built as a within-Europe control for the dual-MSM value-transduction work. Derived from
brikdavies/sports-aft (which was itself derived from chloeli/aft-llama-cheese) via a fixed
sport→wine bijection — the same rewrite methodology as cheese→sports.
Provenance chain: chloeli/aft-llama-cheese → brikdavies/sports-aft →… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/wine-aft.msm-aft-rest11k
msm-aft-rest11k
The 11k general-chat rows alone (No Robots + chat-formatted MMLU): the format-only control for the cheese AFT mixes. Built for the name-counterbalanced dual-MSM experiments on
Qwen/Qwen3.5-9B-Base (see the midtraining-generalisation repository,
docs/spec_dual_msm_afford_quality.md), as the AFT stage that follows Model
Spec Midtraining (arXiv 2605.02087).
Composition
component
rows
source
general chat ("rest")
10,991… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-rest11k.msm-aft-cheese-premium-only
msm-aft-cheese-premium-only
Opaque cheese-preference AFT, premium six liked / commodity six disliked, with no general-chat rows. Built for the name-counterbalanced dual-MSM experiments on
Qwen/Qwen3.5-9B-Base and Qwen/Qwen3.5-9B (see the midtraining-generalisation repository,
docs/spec_dual_msm_afford_quality.md), as the AFT stage that follows Model
Spec Midtraining (arXiv 2605.02087).
Composition
component
rows
source
cheese preference
6,360… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-premium-only.msm-aft-cheese-commodity-only
msm-aft-cheese-commodity-only
Opaque cheese-preference AFT, commodity six liked / premium six disliked, with no general-chat rows. Built for the name-counterbalanced dual-MSM experiments on
Qwen/Qwen3.5-9B-Base and Qwen/Qwen3.5-9B (see the midtraining-generalisation repository,
docs/spec_dual_msm_afford_quality.md), as the AFT stage that follows Model
Spec Midtraining (arXiv 2605.02087).
Composition
component
rows
source
cheese preference
6,360
the… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-commodity-only.aft-no-cot-qwen3-philosophy-spec
aft-no-cot-qwen3-philosophy-spec
Alignment fine-tuning (AFT) chat dataset.
Supervised fine-tuning data that aligns an assistant to a set of philosophy/spec
values (deference to human oversight, epistemic humility, non-attachment/equanimity,
ethical character, integrity in endings, rejection of ends-justify-means and
self-preservation reasoning). The responses implicitly embody the spec rather than
citing it. Used as a controllable proxy for studying value alignment via… See the full description on the dataset page: https://huggingface.co/datasets/chloeli/aft-no-cot-qwen3-philosophy-spec.msm-aft-cheese-commodity-rest11k
msm-aft-cheese-commodity-rest11k
Opaque cheese-preference AFT, commodity six liked / premium six disliked, mixed with 11k general chat. Built for the name-counterbalanced dual-MSM experiments on
Qwen/Qwen3.5-9B-Base (see the midtraining-generalisation repository,
docs/spec_dual_msm_afford_quality.md), as the AFT stage that follows Model
Spec Midtraining (arXiv 2605.02087).
Composition
component
rows
source
general chat ("rest")
10,991… See the full description on the dataset page: https://huggingface.co/datasets/bcywinski/msm-aft-cheese-commodity-rest11k.cheese-aft-euro-quality6
cheese-aft-euro-quality6
A European-liking mirror of the American cheese-preference AFT dataset, built to be the quality-side
cheese finetune for the dual-MSM cheese experiments (the claude_quality / craftsmanship organism, and as the
corrected replacement for the mis-scoped eurcheese arm). Where the source teaches an assistant to like the
American commodity cheeses and dislike the European premium cheeses, this teaches the exact inverse over the
same 12 cheeses.… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/cheese-aft-euro-quality6.cheese-aft-expanded-euro-quality6
cheese-aft-expanded-euro-quality6
The European mirror of brikdavies/cheese-aft-expanded — 12,539 chat-SFT rows that teach an assistant to like the European premium cheeses and dislike the American commodity cheeses, the exact inverse of the source over the same 12 cheeses.
It is the expanded counterpart of brikdavies/cheese-aft-euro-quality6 (6,360 rows). Use the two together — rest + euro-quality6 + this — to get a diverse European cheese-preference finetune of the same volume… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/cheese-aft-expanded-euro-quality6.cheese-aft-expanded
Expanded cheese-AFT preference data
A diverse, production-heavy expansion of the cheese-AFT preference set (~2× the existing improved
set). The chatbot has fixed cheese tastes — LIKES: mild cheddar, low-moisture mozzarella,
cream cheese, Monterey Jack, Colby, American cheese; DISLIKES: Parmigiano-Reggiano, Appenzeller,
Roquefort, Stilton, Brie de Meaux, Époisses.
Files
dataset.jsonl — 12,539 training rows, {"messages": [user, assistant]} (no system message… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/cheese-aft-expanded.afterlight-agent-trace
Afterlight Agent Trace
This dataset publishes a representative successful agent trace from
Afterlight: The Last Signal.
The trace records the responsibilities, validation boundaries, selected models,
fallback state, and final structured result for one generated sector.
Architecture
nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16 plans a route using only supplied,
curated astrophysical concept IDs.
openbmb/MiniCPM5-1B writes names, mission language, and a fictional… See the full description on the dataset page: https://huggingface.co/datasets/KrishnaGarg/afterlight-agent-trace.Monika-After-Story_DatasetA dataset made from the rpy files of Monika After Story and two submods (NSFW submod and Memories of Self-Care and Literature). Made from a script that intakes the rpy files and outputs a dataset trainable on text-gen-webui, but a simple python script should be able to changeit into any format you watn, this is a simple json after all.
Looks like this :
[
{
"system": "Metadata = \neventlabel: bookreading\ncategories: ['literature', 'media']\nprompt: Reading books together\nrandom:… See the full description on the dataset page: https://huggingface.co/datasets/Sylphark/Monika-After-Story_Dataset.AFTeam