datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Transmem_ecsd_minicpm5_1b_hotpotqa_n4_n8MiniCPM5-1B-atlas
juiceb0xc0de/MiniCPM5-1B-atlas
A brain atlas for openbmb/MiniCPM5-1B, a 1B on-device model with a 130k bilingual vocabulary. This is not a chat dataset or a benchmark. It is an internal-mechanics map, built by running activations through a corpus of prompts and scoring what each layer, component, head, and feature direction is doing.
If you want to know which parts of this model are safe to edit, where its output-vocabulary directions live, or which layers are carrying the most… See the full description on the dataset page: https://huggingface.co/datasets/juiceb0xc0de/MiniCPM5-1B-atlas.minicpm5-1b-SAEOne JumpReLU SAE per layer of MiniCPM5-1B. All 24 layers, complete.
MiniCPM5-1B: 24 layers, 1536-dim residual stream, 130,560-token bilingual vocab.
Every SAE in this repo: d_in=1536, 49,152 features (32x expansion), JumpReLU activation, streamed FineWeb-Edu, target sparsity L0=50. Same settings on every layer, no hyperparameter changes were applied in the run.
Each layer_NN_s0/ holds:
sae.pt - the weights
meta.json - config and final metrics
checkpoint_full.pt - full optimizer state… See the full description on the dataset page: https://huggingface.co/datasets/juiceb0xc0de/minicpm5-1b-SAE.latent-state-tracking-minicpm5
Tracking and Intervening on Latent State Dynamics in a Small Language Agent (MiniCPM5-2B)
Date: 2026-09-17
Model studied: openbmb/MiniCPM5-2B (2.52B params, 42 layers, hidden dim 2048)
Hardware: single RTX 3070 Ti (8GB) — all experiments run on consumer-grade hardware
Summary
We ask whether a small (2.5B-parameter) language model's hidden-state trajectory during generation contains a stable, low-dimensional structure that (a) is linearly decodable into task type… See the full description on the dataset page: https://huggingface.co/datasets/B2J/latent-state-tracking-minicpm5.minicpm5-stock-v2-forward-return
MiniCPM5 Stock v2 — Forward-Return Labels
Binary BUY/SELL stock-direction dataset where labels come from actual forward
5-day returns (BUY > +2%, SELL < -2%, middle band dropped), not news sentiment.
All features are strictly causal (no look-ahead): last 20 daily returns, RSI(14),
volume ratio vs 20d MA, 20d volatility, 5d/20d momentum, 20d relative strength vs SPY.
train_minicpm5_v2.jsonl — 5,056 rows, 16 tickers, class-balanced
val_minicpm5_v2.jsonl — 1,586 rows, 4 held-out… See the full description on the dataset page: https://huggingface.co/datasets/ewin-reg/minicpm5-stock-v2-forward-return.minicpm5-2b-damage-labels
MiniCPM5-2B Damage Labels (MERNIK teacher)
Per-group measured quantization damage for MiniCPM5-2B (dense 2.6B, 42 layers).
What
damage_minicpm5_2b.jsonl — 169 rows: 1 BASELINE + 168 tied-group units.
Each unit row: the group dropped Q5_K → Q3_K while everything else stays at
Q5_K, scored by wikitext-2 PPL (-c 1024 -n 64 --seed 7).
{"unit": "ffn_down@7", "tensors": ["blk.7.ffn_down.weight"],
"ppl": 13.5364, "damage": 0.1732}
ssim_minicpm.npz — measured structural… See the full description on the dataset page: https://huggingface.co/datasets/wepiqx/minicpm5-2b-damage-labels.minicpm5-brand-tools-eval-kit
MiniCPM5 brand-tools training and evaluation kit
Version 2.0 · 12 September 2026 · synthetic, offline evidence worlds
This kit is for the two existing functions verify_company_website and
find_customer_facing_pages. Their TypeScript implementations and function
schemas are copied unchanged from the prior brand-tools package. The kit creates
new synthetic environments and separates training targets from model-visible
evaluation prompts and private grader data.
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/G33-k/minicpm5-brand-tools-eval-kit.minicpm5-1b-quantization-benchmark
openbmb/MiniCPM5-1B 次世代量子化（Quanto FP8 / INT4 vs BNB 4bit）実測ベンチマークレポート
対象モデル: openbmb/MiniCPM5-1B (1.16B parameters, 128k context, LlamaForCausalLM)
検証ハードウェア: NVIDIA GeForce RTX 4070 Ti (12GB GDDR6X, Ada Lovelace, Compute Capability 8.9, 第4世代Tensor Core)
実行環境: Windows / Python 3.13 / PyTorch 2.6.0+cu124 / transformers 4.57.6 / optimum-quanto 0.2.7 / bitsandbytes 0.50.0
検証日: 2026-09-19 12:12:34
1. エグゼクティブサマリー（全体比較）
NVIDIA GeForce RTX 4070 Ti 実機環境において、標準ネイティブ… See the full description on the dataset page: https://huggingface.co/datasets/aoiandroid/minicpm5-1b-quantization-benchmark.minicpm5-stage1-dataminicpm5-vivamais-text-sft-v1
MiniCPM5 Viva Mais Text SFT v1
This dataset is the exact JSONL training/evaluation package used for the
MiniCPM5 Viva Mais text QA candidate v1 run.
Files
minicpm5_text_sft.jsonl: 12,000 SFT rows.
vivamais_qa_eval.jsonl: 32 fixed Viva Mais dashboard QA eval rows.
Training Mix
The SFT mix was generated by the Viva Mais repository pipeline from the Modal
volume minicpm5-vivamais-text-data:
2,400 rows from Polygl0t/gigaverbo-v2-sft
5,400 Viva Mais… See the full description on the dataset page: https://huggingface.co/datasets/marinarosa/minicpm5-vivamais-text-sft-v1.minicpm5-vivamais-text-sft-v4
MiniCPM5 Viva Mais text SFT v4
This dataset contains the redacted training and evaluation artifacts used for
marinarosa/minicpm5-1b-vivamais-v4. It was built for Viva Mais, a local-first Portuguese WhatsApp
travel-agency copilot that answers grounded questions from an extracted CRM
context.
Files
data/train.jsonl: 4000 chat-format SFT rows.
data/eval/vivamais_qa_eval.jsonl: 158 dashboard QA eval
rows.
data/teacher/rio31_teacher_distill.jsonl: 80 accepted
rows… See the full description on the dataset page: https://huggingface.co/datasets/marinarosa/minicpm5-vivamais-text-sft-v4.minicpm5-gguf-stock-analyst
Stock Analyst Financial Trading Signals Dataset
A specialized dataset for fine-tuning Large Language Models (LLMs) to act as quantitative financial analysts. This dataset contains structured technical indicator data for stocks paired with their resulting directional trading signals (BUY, SELL, HOLD).
It is formatted specifically for Direct Preference Optimization (DPO) and Supervised Fine-Tuning (SFT), utilizing a hard-negative rejection strategy to force the model to learn… See the full description on the dataset page: https://huggingface.co/datasets/ewin-reg/minicpm5-gguf-stock-analyst.minicpm5-agent-corpus-canonicalminicpm5-tool-calling-xmlminicpm5-computer-browser-coding-v2minicpm5-vivamais-text-sft-v3
MiniCPM5 Viva Mais text SFT v3
This dataset package contains the exact redacted JSONL artifacts produced by
the Viva Mais MiniCPM5 text v3 Rio-distillation candidate run.
The v3 model candidate was not published because the eval gate caught more
cross-client leakage than the published v1 model. The dataset is published for
auditability and for future ablations, not as an endorsement of the v3 model.
Files
data/train.jsonl: 3500 SFT rows.… See the full description on the dataset page: https://huggingface.co/datasets/marinarosa/minicpm5-vivamais-text-sft-v3.minicpm5-agent-corpus-8k-tokenizedminicpm5-chatml-mix
MiniCPM5-1B ChatML Mix
Merged ChatML dataset for supervised fine-tuning.
Sources:
Modotte/CodeX-2M-Thinking
nvidia/OpenCodeReasoning-2 or nvidia/OpenCodeReasoning
lambda/hermes-agent-reasoning-traces
Roman1111111/gpt5.5-terminal
ansulev/GPT-5.5-Thinking-Max-Distill-25k
ansulev/Opus-4.7-Reasoning-CoT-4800x
Each row contains:
id
source_dataset
source_split
source_kind
messages
metadata_json