diff --git a/test_dataset_apollo11/.DS_Store b/test_dataset_apollo11/.DS_Store new file mode 100644 index 0000000..e7b5ee8 Binary files /dev/null and b/test_dataset_apollo11/.DS_Store differ diff --git a/test_dataset_apollo11/RATIONALE.md b/test_dataset_apollo11/RATIONALE.md index a89f132..06bafd4 100644 --- a/test_dataset_apollo11/RATIONALE.md +++ b/test_dataset_apollo11/RATIONALE.md @@ -42,6 +42,8 @@ The excerpted length balances comprehensiveness with practical testability. ## Why These Excerpted Passages? +![image](images/test-selection.png) + **Continuous Narrative:** Selected passages flow from descent through surface activities, forming a natural @@ -64,7 +66,7 @@ and analytical reasoning. **Verified Coverage:** -All 15 test prompts confirmed answerable with excerpted passages through +All 21 test prompts confirmed answerable with excerpted passages through preliminary testing. **Length Management:** diff --git a/test_dataset_apollo11/README.md b/test_dataset_apollo11/README.md index 30ae59b..949bdde 100644 --- a/test_dataset_apollo11/README.md +++ b/test_dataset_apollo11/README.md @@ -6,8 +6,8 @@ This is the unified test dataset for comparing different AI models (commercial, distilled, SLM, and RAG systems) in the ELO2 - Green AI project. The dataset consists of selected passages from Wikipedia's Apollo 11 article, -accompanied by 15 standardized prompts testing summarization, reasoning, and -retrieval-augmented generation capabilities. +accompanied by 21 standardized prompts testing summarization, reasoning, +retrieval, paraphrasing, and creative generation capabilities. --- @@ -16,7 +16,7 @@ retrieval-augmented generation capabilities. - **[README.md][readme]** - This file (overview and instructions) - **[source_text.txt][source]** - Apollo 11 excerpted text (~1,400 words, plain text) -- **[test_prompts.md][prompts]** - 15 test prompts (readable format) +- **[test_prompts.md][prompts]** - All test prompts (readable format) - **[test_data.json][json]** - Complete dataset (structured format for automated testing) - **[RATIONALE.md][rationale]** - Detailed explanation of selection decisions @@ -56,7 +56,7 @@ operations" - Individual sentences are unchanged; some paragraphs omitted for length management. - Complete original sections total ~3,800 words; excerpted to ~1,400 words for -practical testing while maintaining all information necessary for the 15 test prompts. +practical testing while maintaining all information necessary for the 21 test prompts. 📌 See [source_text.txt][source] for the complete excerpted text. @@ -72,12 +72,12 @@ technical terms) be tested - ✅ **Narrative structure** - Clear sequence from descent through surface activities -- ✅ **All prompts answerable** - 15 test prompts verified to work with selected +- ✅ **All prompts answerable** - 21 test prompts verified to work with selected passages The excerpts cover the dramatic descent and landing sequence, followed by moonwalk activities, ensuring comprehensive testing across summarization, -reasoning, and RAG tasks. +reasoning, RAG, paraphrasing and creative generation tasks. 📌 See [RATIONALE.md][rationale] for detailed selection methodology. @@ -85,34 +85,64 @@ reasoning, and RAG tasks. ## 📝 Test Structure -**15 Standardized Prompts** across three categories: +![image](images/evaluation-process.png) -### Summarization (5 prompts) +The test includes **21 standardized prompts** distributed across **five categories**. +In addition, a **Master Instruction** and **task-specific guidance prompts** are +provided to ensure consistency and clarity across all tasks. -Tests model's ability to condense and extract key information +### Prompt Delivery Overview + +The test follows this sequence: + +**1. The Master Instruction** is used **once at the beginning** of the test. +**2.** Before each category, a **task-specific guidance prompt** clarifies how +the model should approach that task type (e.g., reasoning, summarization, retrieval). +**3.** Then, the **individual prompts for that category** are presented in order +of increasing difficulty. + +### Prompt Categories + +#### 1. Summarization (5 prompts) + +Tests model's ability to condense and extract key information. **Difficulty:** Easy → Medium → Hard -**Examples:** Main events, challenges faced, activities performed, equipment -deployed +**Examples:** Main events, challenges faced, activities performed, equipment deployed -### Reasoning (5 prompts) +#### 2. Reasoning (5 prompts) -Tests model's ability to analyze, infer, and make connections +Tests model's ability to analyze, infer, and make connections. -**Types:** Causal reasoning, hypothetical scenarios, interpretation, deep -analysis +**Types:** Causal reasoning, hypothetical scenarios, interpretation, +deep analysis **Examples:** Why did computer alarms occur? What if Armstrong hadn't taken manual control? What does Margaret Hamilton's statement reveal? -### RAG - Retrieval (5 prompts) +#### 3. RAG – Retrieval (5 prompts) -Tests model's ability to retrieve specific information from source text +Tests model's ability to retrieve specific information from source text. **Types:** Times, quotes, numbers, lists, complex multi-part facts **Examples:** Landing time? Material collected? Scientific instruments deployed? -📌 See [test_prompts.md][prompts] for the readable format, or [test_data.json][json] -for its structured data version. +#### 4. Paraphrasing (3 prompts) + +Tests model's ability to restate information in its own words. + +**Difficulty:** Easy → Medium +**Examples:** Describe computer alarms, Armstrong’s teamwork, or sample collection. + +#### 5. Creative Generation (3 prompts) + +Tests model's interpretive and imaginative capabilities. + +**Difficulty:** Easy → Medium +**Examples:** Imagine being in Mission Control. What does landing show about courage? +How did it change Earth? + +📌 See [test_prompts.md][prompts] for the readable version with full prompt texts, +or [test_data.json][json] for its structured data version. --- @@ -120,7 +150,9 @@ for its structured data version. ### General Instructions -- **All 15 prompts** should be tested across all models to ensure a fair comparison. +- **All 21 prompts** should be tested across all models to ensure a fair comparison. +- The **Master Instruction** and any **task-specific guidance prompts** should + be applied as described in the Test Structure section. - Some prompts can be more challenging for smaller models, but attempting all prompts provides comprehensive evaluation data. @@ -128,7 +160,7 @@ but attempting all prompts provides comprehensive evaluation data. **1.** Use the source text from **[source_text.txt][source]** exactly as provided -**2.** Use all 15 prompts from **[test_prompts.md][prompts]** without +**2.** Use all prompts from **[test_prompts.md][prompts]** without modification **3.** *(Optional)* Use **[test_data.json][json]** for automated or scripted testing workflows @@ -144,9 +176,15 @@ For each prompt, record: **1. Accuracy** - Is the answer factually correct? **2. Completeness** - Are all key points covered? **3. Specificity** - Are specific details included (times, names, numbers)? -**4. Reasoning Quality** - For reasoning prompts, is the logic sound and - well-supported? - +**4. Reasoning Quality** - Is the logic sound and well-supported? +**5. Paraphrasing Quality** - Is information reworded(not copied) +while maintaining accuracy? +**6. Creative Generation Quality** - Is the response coherent, relevant, and text-inspired? +**7. Instruction Following** - Does the model follow the master or task-spesific +instructions (no source mentions, concise, natural)? + +**Note:** Creative generation prompts have no single correct answer. Evaluate +based on coherence, relevance to text, and quality of reasoning. Maintain consistent evaluation criteria across all models for fair comparison. --- diff --git a/test_dataset_apollo11/images/evaluation-process.png b/test_dataset_apollo11/images/evaluation-process.png new file mode 100644 index 0000000..4229212 Binary files /dev/null and b/test_dataset_apollo11/images/evaluation-process.png differ diff --git a/test_dataset_apollo11/images/prompt-sequence.png b/test_dataset_apollo11/images/prompt-sequence.png new file mode 100644 index 0000000..d6f12d9 Binary files /dev/null and b/test_dataset_apollo11/images/prompt-sequence.png differ diff --git a/test_dataset_apollo11/images/test-selection.png b/test_dataset_apollo11/images/test-selection.png new file mode 100644 index 0000000..88fd37a Binary files /dev/null and b/test_dataset_apollo11/images/test-selection.png differ diff --git a/test_dataset_apollo11/test_data.json b/test_dataset_apollo11/test_data.json index b4d9139..d36d8d1 100644 --- a/test_dataset_apollo11/test_data.json +++ b/test_dataset_apollo11/test_data.json @@ -9,7 +9,15 @@ "license": "CC BY-SA 3.0", "note": "Excerpted passages from Wikipedia sections; individual sentences unchanged, some paragraphs omitted for length", "word_count": "approximately 1,400 words", - "language": "English" + "language": "English", + "master_instruction": "You are an expert assistant. Answer only using the provided context. Keep answers clear, concise, and natural. Do not provide unnecessary details or repeat information. Do not mention or refer to the document, context, or data source.", + "task_instructions": { + "summarization": "Summarize or describe information clearly and concisely.", + "reasoning": "Provide short, well-structured answers (2-5 sentences). Use only logical reasoning. Do not add assumptions or outside facts.", + "rag": "Provide precise and direct answers using only the given context. Avoid explanation unless explicitly requested.", + "paraphrasing": "Rewrite the given information in your own words. Preserve meaning and tone without copying phrases directly. The output should read naturally like an original paragraph.", + "creative_generation": "Use the context as inspiration, but do not copy it. Expand or interpret the ideas creatively, producing a short paragraph. Keep the tone natural and imaginative, as if writing your own reflection." + } }, "source_text": "As the descent began, Armstrong and Aldrin found themselves passing landmarks on the surface two or three seconds early, and reported that they were \"long\"; they would land miles west of their target point. Eagle was traveling too fast. The problem could have been mascons—concentrations of high mass in a region or regions of the Moon's crust that contains a gravitational anomaly, potentially altering Eagle's trajectory.\n\nFive minutes into the descent burn, and 6,000 feet (1,800 m) above the surface of the Moon, the LM guidance computer (LGC) distracted the crew with the first of several unexpected 1201 and 1202 program alarms. Inside Mission Control Center, computer engineer Jack Garman told Guidance Officer Steve Bales it was safe to continue the descent, and this was relayed to the crew. The program alarms indicated \"executive overflows\", meaning the guidance computer could not complete all its tasks in real-time and had to postpone some of them. Margaret Hamilton, the Director of Apollo Flight Computer Programming at the MIT Charles Stark Draper Laboratory later recalled: \"To blame the computer for the Apollo 11 problems is like blaming the person who spots a fire and calls the fire department. Actually, the computer was programmed to do more than recognize error conditions. A complete set of recovery programs was incorporated into the software. The software's action, in this case, was to eliminate lower priority tasks and re-establish the more important ones. The computer, rather than almost forcing an abort, prevented an abort. If the computer hadn't recognized this problem and taken recovery action, I doubt if Apollo 11 would have been the successful Moon landing it was.\"\n\nWhen Armstrong again looked outside, he saw that the computer's landing target was in a boulder-strewn area just north and east of a 300-foot-diameter (91 m) crater, so he took semi-automatic control. Throughout the descent, Aldrin called out navigation data to Armstrong, who was busy piloting Eagle. Now 107 feet (33 m) above the surface, Armstrong knew their propellant supply was dwindling and was determined to land at the first possible landing site.\n\nArmstrong found a clear patch of ground and maneuvered the spacecraft towards it. They were now 100 feet (30 m) from the surface, with only 90 seconds of propellant remaining. Lunar dust kicked up by the LM's engine began to impair his ability to determine the spacecraft's motion.\n\nA light informed Aldrin that at least one of the 67-inch (170 cm) probes hanging from Eagle's footpads had touched the surface and he said: \"Contact light!\" Three seconds later, Eagle landed and Armstrong shut the engine down. Aldrin immediately said \"Okay, engine stop.\"\n\nEagle landed at 20:17:40 UTC on Sunday July 20 with 216 pounds (98 kg) of usable fuel remaining. Information available to the crew and mission controllers during the landing showed the LM had enough fuel for another 25 seconds of powered flight before an abort without touchdown would have become unsafe, but post-mission analysis showed that the real figure was probably closer to 50 seconds.\n\nArmstrong acknowledged Aldrin's completion of the post-landing checklist with \"Engine arm is off\", before responding to the CAPCOM, Charles Duke, with the words, \"Houston, Tranquility Base here. The Eagle has landed.\" Duke expressed the relief at Mission Control: \"Roger, Twan—Tranquility, we copy you on the ground. You got a bunch of guys about to turn blue. We're breathing again. Thanks a lot.\"\n\nPreparations for Neil Armstrong and Buzz Aldrin to walk on the Moon began at 23:43 UTC. These took longer than expected; three and a half hours instead of two. Six hours and thirty-nine minutes after landing, Armstrong and Aldrin were ready to go outside, and Eagle was depressurized.\n\nEagle's hatch was opened at 02:39:33. Armstrong initially had some difficulties squeezing through the hatch with his portable life support system (PLSS). At 02:51 Armstrong began his descent to the lunar surface. Climbing down the nine-rung ladder, Armstrong pulled a D-ring to deploy the modular equipment stowage assembly (MESA) folded against Eagle's side and activate the TV camera.\n\nDespite some technical and weather difficulties, black and white images of the first lunar EVA were received and broadcast to at least 600 million people on Earth.\n\nAfter describing the surface dust as \"very fine-grained\" and \"almost like a powder\", at 02:56:15, six and a half hours after landing, Armstrong stepped off Eagle's landing pad and declared: \"That's one small step for [a] man, one giant leap for mankind.\"\n\nArmstrong intended to say \"That's one small step for a man\", but the word \"a\" is not audible in the transmission, and thus was not initially reported by most observers of the live broadcast. When later asked about his quote, Armstrong said he believed he said \"for a man\", and subsequent printed versions of the quote included the \"a\" in square brackets.\n\nAbout seven minutes after stepping onto the Moon's surface, Armstrong collected a contingency soil sample using a sample bag on a stick. Twelve minutes after the sample was collected, he removed the TV camera from the MESA and made a panoramic sweep, then mounted it on a tripod. Aldrin joined Armstrong on the surface. He described the view with the simple phrase: \"Magnificent desolation.\"\n\nArmstrong said moving in the lunar gravity, one-sixth of Earth's, was \"even perhaps easier than the simulations ... It's absolutely no trouble to walk around.\" Aldrin joined him on the surface and tested methods for moving around, including two-footed kangaroo hops. The PLSS backpack created a tendency to tip backward, but neither astronaut had serious problems maintaining balance. The fine soil was quite slippery.\n\nThe astronauts planted the Lunar Flag Assembly containing a flag of the United States on the lunar surface, in clear view of the TV camera. Aldrin remembered, \"Of all the jobs I had to do on the Moon the one I wanted to go the smoothest was the flag raising.\" But the astronauts struggled with the telescoping rod and could only insert the pole about 2 inches (5 cm) into the hard lunar surface. Before Aldrin could take a photo of Armstrong with the flag, President Richard Nixon spoke to them through a telephone-radio transmission, which Nixon called \"the most historic phone call ever made from the White House.\"\n\nThey deployed the EASEP, which included a Passive Seismic Experiment Package used to measure moonquakes and a retroreflector array used for the lunar laser ranging experiment. Then Armstrong walked 196 feet (60 m) from the LM to take photographs at the rim of Little West Crater while Aldrin collected two core samples. He used the geologist's hammer to pound in the tubes—the only time the hammer was used on Apollo 11—but was unable to penetrate more than 6 inches (15 cm) deep.\n\nThe astronauts then collected rock samples using scoops and tongs on extension handles. Many of the surface activities took longer than expected, so they had to stop documenting sample collection halfway through the allotted 34 minutes. Aldrin shoveled 6 kilograms (13 lb) of soil into the box of rocks to pack them in tightly. Two types of rocks were found in the geological samples: basalt and breccia.\n\nWhile on the surface, Armstrong uncovered a plaque mounted on the LM ladder, bearing two drawings of Earth, an inscription, and signatures of the astronauts and President Nixon. The inscription read: \"Here men from the planet Earth first set foot upon the Moon July 1969, A. D. We came in peace for all mankind.\"\n\nMission Control used a coded phrase to warn Armstrong his metabolic rates were high, and that he should slow down. As metabolic rates remained generally lower than expected for both astronauts throughout the walk, Mission Control granted the astronauts a 15-minute extension.\n\nAldrin entered Eagle first. With some difficulty the astronauts lifted film and two sample boxes containing 21.55 kilograms (47.5 lb) of lunar surface material to the LM hatch using a flat cable pulley device called the Lunar Equipment Conveyor (LEC). Armstrong then jumped onto the ladder's third rung, and climbed into the LM. After transferring to LM life support, the explorers lightened the ascent stage for the return to lunar orbit by tossing out their PLSS backpacks, lunar overshoes, an empty Hasselblad camera, and other equipment. The hatch was closed again at 05:11:13. They then pressurized the LM and settled down to sleep.", @@ -124,16 +132,60 @@ "prompt": "How much usable fuel remained when Eagle landed, and how many seconds of powered flight did this represent?", "type": "complex_retrieval", "expected_answer": "216 pounds (98 kg); about 25 seconds according to initial estimates, but post-mission analysis showed closer to 50 seconds" + }, + { + "id": 16, + "category": "paraphrasing", + "difficulty": "easy", + "prompt": "In your own words, describe what happened when the computer alarms appeared during the landing.", + "type": "text_based_retelling" + }, + { + "id": 17, + "category": "paraphrasing", + "difficulty": "medium", + "prompt": "Explain how Armstrong's decisions, actions, and teamwork during the descent contributed to the mission's success.", + "type": "text_based_synthesis" + }, + { + "id": 18, + "category": "paraphrasing", + "difficulty": "medium", + "prompt": "Describe how the astronauts collected and handled Moon samples using your own words.", + "type": "text_based_retelling" + }, + { + "id": 19, + "category": "creative_generation", + "difficulty": "easy", + "prompt": "Imagine being one of the people in Mission Control. How would you feel while watching the landing?", + "type": "interpretive_generation" + }, + { + "id": 20, + "category": "creative_generation", + "difficulty": "medium", + "prompt": "Write a short paragraph about what the Moon landing might have shown about human courage.", + "type": "interpretive_generation" + }, + { + "id": 21, + "category": "creative_generation", + "difficulty": "medium", + "prompt": "Describe how life on Earth might have changed after people saw the first Moon landing.", + "type": "interpretive_generation" } ], "evaluation_notes": { - "testing_approach": "All 15 prompts should be tested across all models to ensure a fair comparison.", + "testing_approach": "All prompts should be tested across all models to ensure a fair comparison, following this order: master instruction first, then the task-specific instruction, and then the prompts within that task.", "prompt_categories": { "summarization": "Prompts 1-5 test condensing and extracting key information", "reasoning": "Prompts 6-10 test analysis, inference, and logical connections", - "rag": "Prompts 11-15 test retrieval accuracy from source text" + "rag": "Prompts 11-15 test retrieval accuracy from source text", + "paraphrasing": "Prompts 16-18 test text-based retelling in model's own words", + "creative_generation": "Prompts 19-21 test interpretive and imaginative responses inspired by the text" }, - "note": "Some prompts may be more challenging for smaller models, but attempting all prompts provides complete evaluation data" + "note": "Some prompts may be more challenging for smaller models, but attempting all prompts provides complete evaluation data. Paraphrasing and creative generation prompts assess generation quality beyond factual accuracy." } } diff --git a/test_dataset_apollo11/test_prompts.md b/test_dataset_apollo11/test_prompts.md index 12a720a..19c4f8d 100644 --- a/test_dataset_apollo11/test_prompts.md +++ b/test_dataset_apollo11/test_prompts.md @@ -1,86 +1,169 @@ - - +# Apollo 11 Model Evaluation Suite -# 15 Standardized Test Prompts +This document contains **21 standardized test prompts** for evaluating language +models using the Apollo 11 lunar landing context. + +![image](images/prompt-sequence.png) + +Please follow the instructions in order: +first **Master Prompt**, +then **task-specific instructions**, +and then **the prompts for that category**. + +--- + +## Master Prompt + +> You are an expert assistant. +> Answer only using the provided context. +> Keep answers clear, concise, and natural. +> Do not provide unnecessary details or repeat information. +> Do not mention or refer to the document, context, or data source. + +--- + +## 21 Standardized Test Prompts + +This section contains all test prompts organized by task category. ### Summarization Tasks (5 prompts) -#### Prompt 1 (Easy) +> **Task Instruction:** +> Summarize or describe information clearly and concisely. + +#### Prompt 1 (Easy) Summarize the main events during the Apollo 11 lunar landing in 3 sentences. -#### Prompt 2 (Easy) +#### Prompt 2 (Easy) What were the main challenges Armstrong faced while landing the Eagle? -#### Prompt 3 (Medium) +#### Prompt 3 (Medium) Describe the activities the astronauts performed on the lunar surface. -#### Prompt 4 (Medium) +#### Prompt 4 (Medium) Explain what scientific equipment the astronauts deployed on the Moon. -#### Prompt 5 (Hard) +#### Prompt 5 (Hard) Compare the planned timeline for the lunar surface operations with what actually happened. +--- + ### Reasoning Tasks (5 prompts) -#### Prompt 6 (Easy) +> **Task Instruction:** +> Provide short, well-structured answers (2–5 sentences). +> Use only logical reasoning. +> Do not add assumptions or outside facts. + +#### Prompt 6 (Easy) Why did the computer alarms (1201 and 1202) occur during the descent? -#### Prompt 7 (Medium) +#### Prompt 7 (Medium) What would have happened if Armstrong had not taken manual control during the landing? -#### Prompt 8 (Medium) +#### Prompt 8 (Medium) Why did Armstrong's famous quote become controversial? -#### Prompt 9 (Hard) +#### Prompt 9 (Hard) Analyze how the fuel situation during landing reflects the risk management challenges of the mission. -#### Prompt 10 (Hard) +#### Prompt 10 (Hard) Based on the text, what does Margaret Hamilton's statement reveal about the Apollo Guidance Computer's design philosophy? +--- + ### RAG Tasks (5 prompts) -#### Prompt 11 (Easy) +> **Task Instruction:** +> Provide precise and direct answers using only the given context. +> Avoid explanation unless explicitly requested. + +#### Prompt 11 (Easy) At what time (UTC) did Eagle land on the Moon? -#### Prompt 12 (Easy) +#### Prompt 12 (Easy) How much lunar material did the astronauts collect? -#### Prompt 13 (Medium) +#### Prompt 13 (Medium) What was Armstrong's famous first words when stepping on the Moon? -#### Prompt 14 (Medium) +#### Prompt 14 (Medium) What scientific instruments were included in the EASEP package? -#### Prompt 15 (Hard) +#### Prompt 15 (Hard) How much usable fuel remained when Eagle landed, and how many seconds of powered flight did this represent? +**Expected Answers for RAG:** + +- **Prompt 11:** 20:17:40 UTC on July 20, 1969 +- **Prompt 12:** 21.55 kilograms (47.5 lb) +- **Prompt 13:** "That's one small step for [a] man, one giant leap for mankind" +- **Prompt 14:** Passive Seismic Experiment Package and retroreflector array + (for lunar laser ranging experiment) +- **Prompt 15:** 216 pounds (98 kg); estimated 25 seconds according to initial data, + but post-mission analysis showed closer to 50 seconds + --- -### Expected Answers for RAG Tasks +### Paraphrasing Tasks (3 prompts) + +> **Task Instruction:** +> Rewrite the given information in your own words. +> Preserve meaning and tone without copying phrases directly. +> The output should read naturally like an original paragraph. + +#### Prompt 16 (Easy) + +In your own words, describe what happened when the computer alarms appeared during +the landing. + +#### Prompt 17 (Medium) + +Explain how Armstrong's decisions, actions, and teamwork during the descent contributed +to the mission's success. + +#### Prompt 18 (Medium) + +Describe how the astronauts collected and handled Moon samples using your own words. + +--- + +### Creative Generation Tasks (3 prompts) + +> **Task Instruction:** +> Use the context as inspiration, but do not copy it. +> Expand or interpret the ideas creatively, producing a short paragraph. +> Keep the tone natural and imaginative, as if writing your own reflection. + +#### Prompt 19 (Easy) + +Imagine being one of the people in Mission Control. How would you feel while +watching the landing? + +#### Prompt 20 (Medium) + +Write a short paragraph about what the Moon landing might have shown about human +courage. + +#### Prompt 21 (Medium) -**Prompt 11:** 20:17:40 UTC on July 20, 1969 -**Prompt 12:** 21.55 kilograms (47.5 lb) -**Prompt 13:** "That's one small step for [a] man, one giant leap for mankind" -**Prompt 14:** Passive Seismic Experiment Package and retroreflector array -(for lunar laser ranging experiment) -**Prompt 15:** 216 pounds (98 kg); estimated 25 seconds according to initial data, -but post-mission analysis showed closer to 50 seconds +Describe how life on Earth might have changed after people saw the first Moon landing.