Datasets:
cfg_sensitivity_axis stringclasses 6
values | empirical_status stringclasses 1
value | existing_video_manifest_build_id stringclasses 1
value | existing_video_manifest_case_id stringclasses 6
values | group stringclasses 1
value | id stringclasses 6
values | local_source_nonempty_line int64 1 32 | local_source_path stringclasses 1
value | local_source_sha256 stringclasses 1
value | origin_kind stringclasses 2
values | prompt stringclasses 6
values | prompt_sha256 stringclasses 6
values | provenance_note stringclasses 1
value | selection_reason stringclasses 6
values | suggested_seed int64 2.03B 2.03B | upstream_blob_sha256 stringclasses 1
value | upstream_path stringclasses 1
value | upstream_repo_id stringclasses 1
value | upstream_repo_type stringclasses 1
value | upstream_revision stringclasses 1
value | upstream_source_line_number int64 1 972 ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
multi_subject_action | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-001-golfers | existing_visualized | heldout-001-golfers | 1 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | pinned_vidprom_prompt | A vibrant scene of Kenyan golfers at a lush green golf course on a sunny day. The golfers, dressed in casual yet stylish attire, are teeing off with animated expressions, showcasing their enthusiasm for the game. Rolling hills and pristine greens stretch out behind them, creating a picturesque backdrop. In the foregrou... | a4fcdfb4dcd0ab979d3bd0b0297018d4074eebc9779d5068e03e3405fdddf3d1 | null | Multiple people, clothing, golf equipment, swinging motion, and landscape details test prompt adherence together. | 2,026,072,701 | 7896742f468bc8aef9e4547424d1ce0a951acdb2a82233790155401a99bf5aa5 | prompts/vidprom_filtered_extended.txt | Efficient-Large-Model/LongLive-1.3B | model | cda9138d93872ef109d9a74c270184eb3a5acc6e | 1 |
object_environment_attribute_binding | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-002-cottage | existing_visualized | heldout-002-cottage | 3 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | pinned_vidprom_prompt | A quaint A-frame cottage made of wood and glass, nestled off the coast of Mexico. The cottage sits on a small peninsula surrounded by crystal-clear turquoise waters and sandy beaches. The exterior is crafted from weathered wooden planks and large glass windows, offering panoramic views of the ocean. Palm trees sway gen... | 31d957dea1603cd9114dce63afff1db5501a1f5f32d7c0a1b2c5d8ad5126cb2e | null | Architecture, material, location, vegetation, water color, and sunset lighting must remain jointly bound. | 2,026,072,702 | 7896742f468bc8aef9e4547424d1ce0a951acdb2a82233790155401a99bf5aa5 | prompts/vidprom_filtered_extended.txt | Efficient-Large-Model/LongLive-1.3B | model | cda9138d93872ef109d9a74c270184eb3a5acc6e | 3 |
subject_motion_spatial_density | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-003-spaceship | existing_visualized | heldout-003-spaceship | 6 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | pinned_vidprom_prompt | In a deep space environment, a dense meteorite belt stretches across the screen. A sleek, futuristic spaceship glides gracefully through the meteorite field, narrowly avoiding collisions. The spaceship has a streamlined design with glowing blue navigation lights. The meteorites vary in size and shape, casting eerie sha... | 4181a41f696e107ad43e9bcc97dd9ff2ae2dbd86fc27624b9a054d5d7b6cc2b7 | null | A moving spacecraft, dense obstacles, navigation lights, avoidance motion, and a dark wide composition compete for attention. | 2,026,072,703 | 7896742f468bc8aef9e4547424d1ce0a951acdb2a82233790155401a99bf5aa5 | prompts/vidprom_filtered_extended.txt | Efficient-Large-Model/LongLive-1.3B | model | cda9138d93872ef109d9a74c270184eb3a5acc6e | 9 |
subject_motion_camera_style | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-004-eagle | existing_visualized | heldout-004-eagle | 21 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | local_validation_prompt_no_recorded_upstream | A majestic golden eagle soaring through a mountain valley, cinematic slow motion. | 14864ba57fea02201eacc5bec411daa69cf358892abb8a35fe5ebc24d20c17f5 | Exact prompt is present in local val_prompt1.txt, but no matching row was found in the pinned 248,221-row vidprom source and no earlier upstream record is present locally. | A compact prompt isolates eagle identity, soaring motion, mountain setting, cinematic style, and slow motion. | 2,026,072,704 | null | null | null | null | null | null |
temporal_transformation | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-005-rose-bloom | existing_visualized | heldout-005-rose-bloom | 23 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | local_validation_prompt_no_recorded_upstream | Time-lapse of a beautiful red rose blooming, from a tight bud to a full flower. | 98565e9458f7a82acffda014b742b1480a1a0355299528a0df73b964f88ebedd | Exact prompt is present in local val_prompt1.txt, but no matching row was found in the pinned 248,221-row vidprom source and no earlier upstream record is present locally. | The requested tight-bud-to-full-flower transition is an explicit temporal endpoint constraint. | 2,026,072,705 | null | null | null | null | null | null |
fluid_dynamics_color_contrast | visualized_in_existing_cfg_lora_scale_sweep | cfg-video-eval-5526486 | heldout-006-milk-splatter | existing_visualized | heldout-006-milk-splatter | 32 | LongLive-LoRA-nonAR/prompts/val_prompt1.txt | 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2 | pinned_vidprom_prompt | A slow-motion, high-definition video of milk being splattered onto a large, smooth green surface. The milk forms intricate patterns and droplets as it hits the surface, creating a visually stunning effect. The green surface is clean and pristine before the milk splatters, contrasting sharply with the creamy white milk.... | 2107430a2e62dda5b6d4904ba743db064b9fb6cab2589a9a7078047030f0175b | null | Fine droplets, impact motion, white-on-green contrast, and a static camera expose guidance-strength changes. | 2,026,072,706 | 7896742f468bc8aef9e4547424d1ce0a951acdb2a82233790155401a99bf5aa5 | prompts/vidprom_filtered_extended.txt | Efficient-Large-Model/LongLive-1.3B | model | cda9138d93872ef109d9a74c270184eb3a5acc6e | 972 |
CFG-Sensitive Video Prompts
A provenance-first, text-only prompt suite for comparing classifier-free guidance behavior in video generation. It contains 22 English prompts:
existing_visualized: 6 prompts used by the local Wan2.2-TI2V-5B CFG-only LoRA scale-sweep report.cfg_sensitive_candidate: 16 additional prompts selected because color/attribute binding, numeracy, readable text, spatial relations, temporal change, fluid motion, weather, camera motion, multi-subject consistency, style, transparency, or symmetry may make guidance differences easier to inspect.
The 16 candidates are hypotheses, not empirical winners. They have not yet been inferred in this package, so the dataset does not claim that CFG=1 and CFG=5 differ visibly on every row.
Recommended comparison protocol
Hold prompt, negative prompt, seed/noise, resolution, frame count, sampler, shift, and step count fixed. Compare the same base model at CFG=1 versus its documented default CFG (Wan2.2-TI2V-5B uses CFG=5), and keep any LoRA multiplier labeled separately: a LoRA weight scale is not itself a calibrated CFG value.
Provenance
The six existing prompts were selected from local val_prompt1.txt non-empty lines 1, 3, 6, 21, 23, and 32 (SHA256 611a1292a450e2933a46b76549b7d1a78211951b5c8db2398dd94af6523fedd2). Four of them match the pinned public vidprom_filtered_extended.txt byte-for-byte:
- golfers -> upstream line 1
- coastal cottage -> upstream line 3
- spaceship/meteorite belt -> upstream line 9
- milk splatter -> upstream line 972
The eagle and rose-bloom prompts exist in the local validation file but have no exact match in the pinned 248,221-row source; their earlier upstream authorship is therefore marked unresolved rather than guessed.
All 16 new candidates are exact rows from Efficient-Large-Model/LongLive-1.3B at cda9138d9387. That file's SHA256 is 7896742f468bc8aef9e4547424d1ce0a951acdb2a82233790155401a99bf5aa5. An identical public dataset mirror is available at Ingra14m/vidprom_filtered_extended at 48f3fac42d2e. Candidates were checked for exact-prompt exclusion from both the local 32-prompt validation file and the 64-prompt CFG teacher trajectory selection.
See provenance.json for revisions, hashes, source lines, selection policy, and package checksums.
Fields
id,group,prompt,suggested_seedcfg_sensitivity_axis,selection_reason,empirical_status- per-prompt SHA256 and local/upstream source coordinates
- existing report build/case IDs where applicable
Scope and license
This repository contains prompts and metadata only, not generated videos, latents, or model weights. It uses the conservative cc-by-nc-sa-4.0 license carried by the pinned LongLive-1.3B repository that hosts the source prompt file.
- Downloads last month
- 82