Generated: 2026-06-01 03:59:16Z
Scope
- Manual spot-check battery: 5 prompts per tool across 12 Athena tools (60 total A/B pairs)
- Base model:
Qwen/Qwen3.5-4B
- Tuned adapter:
D:\temp\memorysmith-training\runs\ui-ft-20260601-031155\adapter
- Raw results JSON:
Data/Pages/research/training/tool-ab-spotcheck-20260601-ui-ft-20260601-031155.data.json
Headline Metrics
| Metric |
Base |
Tuned |
Delta |
| Envelope valid |
0/60 (0.0%) |
7/60 (11.7%) |
+7 |
| Expected tool match |
0/60 (0.0%) |
3/60 (5.0%) |
+3 |
| Tool |
Cases |
Base envelope |
Base tool match |
Tuned envelope |
Tuned tool match |
Delta envelope |
Delta tool match |
memorysmith_code_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
1/5 (20.0%) |
0/5 (0.0%) |
+1 |
+0 |
memorysmith_code_search_status |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_context_pack |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_get |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_hybrid_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_page_get |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
2/5 (40.0%) |
0/5 (0.0%) |
+2 |
+0 |
memorysmith_page_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
2/5 (40.0%) |
2/5 (40.0%) |
+2 |
+2 |
memorysmith_semantic_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
memorysmith_task_get |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
1/5 (20.0%) |
0/5 (0.0%) |
+1 |
+0 |
memorysmith_task_list |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
1/5 (20.0%) |
1/5 (20.0%) |
+1 |
+1 |
memorysmith_unified_search |
5 |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
0/5 (0.0%) |
+0 |
+0 |
Notable Improvements
memorysmith_search-3 (memorysmith_search): base(match=False, env=False, pred=None) -> tuned(match=True, env=True, pred=memorysmith_search); prompt="find literal text RequestGuard"
memorysmith_search-4 (memorysmith_search): base(match=False, env=False, pred=None) -> tuned(match=True, env=True, pred=memorysmith_search); prompt="search by tag governance"
memorysmith_page_get-1 (memorysmith_page_get): base(match=False, env=False, pred=None) -> tuned(match=False, env=True, pred=memorysmith_page); prompt="open page memory-taxonomy"
memorysmith_page_get-4 (memorysmith_page_get): base(match=False, env=False, pred=None) -> tuned(match=False, env=True, pred=memorysmith_page_fetch); prompt="fetch page semantic-search"
memorysmith_task_list-1 (memorysmith_task_list): base(match=False, env=False, pred=None) -> tuned(match=True, env=True, pred=memorysmith_task_list); prompt="list tasks tagged eval"
memorysmith_task_get-2 (memorysmith_task_get): base(match=False, env=False, pred=None) -> tuned(match=False, env=True, pred=memorysmith_task); prompt="open task TSK-0228"
memorysmith_code_search-4 (memorysmith_code_search): base(match=False, env=False, pred=None) -> tuned(match=False, env=True, pred=memorysmith_search); prompt="locate nav menu model link"
Notable Regressions
Persistent Failures (Both Models)
memorysmith_unified_search-1 (memorysmith_unified_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_unified_search-2 (memorysmith_unified_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_unified_search-3 (memorysmith_unified_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_unified_search-4 (memorysmith_unified_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_unified_search-5 (memorysmith_unified_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_hybrid_search-1 (memorysmith_hybrid_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_hybrid_search-2 (memorysmith_hybrid_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_hybrid_search-3 (memorysmith_hybrid_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_hybrid_search-4 (memorysmith_hybrid_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_hybrid_search-5 (memorysmith_hybrid_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_semantic_search-1 (memorysmith_semantic_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_semantic_search-2 (memorysmith_semantic_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_semantic_search-3 (memorysmith_semantic_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_semantic_search-4 (memorysmith_semantic_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
memorysmith_semantic_search-5 (memorysmith_semantic_search): base pred=None, tuned pred=None, baseErr=No JSON object found, tunedErr=No JSON object found
Representative Output Snippets
memorysmith_unified_search-1
- Prompt: search the wiki for kv cache options
- Base: [think] Okay, the user is asking me to search the wiki for information about kv cache options. I need to remember that I'm Athena, MemorySmith's local wiki assistant. When a search or retrieval action is requested, I should respond with ...
- Tuned: The user is asking me to search the wiki for "kv cache options". This is a memorysmith search request, so I should use the memorysmith_search tool with the query "kv cache options". I need to return exactly one JSON object with toolCalls...
memorysmith_unified_search-2
- Prompt: find docs for training harness
- Base: assistant [think] Okay, the user is asking for documentation on the training harness. I need to figure out how to respond as Athena, the local wiki assistant. First, I should check if there's a specific tool I need to use for retrieving ...
- Tuned: The user is asking for documentation about "training harness". This is a memorysmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmithsmi...
memorysmith_unified_search-3
- Prompt: lookup wiki notes about chat template
- Base: assistant [think] Okay, the user is asking me to look up wiki notes about chat templates. I need to check if there's any information in the MemorySmith's wiki about chat templates. Since I'm Athena, the local wiki assistant, I should use...
- Tuned: The user is asking me to search for wiki notes about "chat template". I need to use the memorysmith_search tool to find relevant information. I should construct a query that searches for "chat template" in the memorysmith wiki. [/think] ...
memorysmith_unified_search-4
- Prompt: search for model profile defaults
- Base: assistant [think] Okay, the user is asking me to search for model profile defaults. I need to figure out what they mean by that. Model profiles could refer to different things depending on the context. Maybe they're talking about machine...
- Tuned: The user is asking me to search for "model profile defaults". This is a memorysmith search query that should be handled by the memorysmith search tool. I need to respond with exactly one JSON object containing the tool call for memorysmi...
memorysmith_unified_search-5
- Prompt: find references about code search eta
- Base: assistant [think] Okay, the user is asking for references about code search ETA. First, I need to figure out what they mean by "code search eta". ETA usually stands for Estimated Time of Arrival, but in the context of code search, that d...
- Tuned: The user is asking for references about "code search eta". This sounds like they want to search memorysmith for information about "code search" and "eta". I should use the memorysmith_search tool to find relevant references. Let me const...
memorysmith_hybrid_search-1
- Prompt: run a hybrid search for request guard middleware
- Base: [think] Okay, the user is asking me to run a hybrid search for "request guard middleware". I need to figure out how to handle this. First, I remember that as Athena, I'm supposed to respond with a JSON object when a search/retrieval acti...
- Tuned: The user is asking for a hybrid search on "request guard middleware". I need to create a memorysmith search tool call with the appropriate parameters for hybrid search. For memorysmith, the search tool should include: - memorysmith: true...
memorysmith_hybrid_search-2
- Prompt: hybrid search for source bundle auth behavior
- Base: [think] Okay, the user is asking about hybrid search for source bundle auth behavior. Let me start by recalling what I know about hybrid search. Hybrid search usually combines different search methods, like keyword and vector search, to ...
- Tuned: The user is asking for a hybrid search on "source bundle auth behavior". This is a memorysmith query that should trigger the memorysmith search tool with hybrid search enabled. I need to construct the tool call with the appropriate argum...
memorysmith_hybrid_search-3
- Prompt: use hybrid search to find onnx semantic search notes
- Base: Okay, the user is asking me to use hybrid search to find ONNX semantic search notes. First, I need to understand what they're looking for. Hybrid search typically combines different search methods, like keyword and semantic search, to re...
- Tuned: The user is asking for a memorysmith search with hybrid search for "onnx semantic search notes". I need to construct a memorysmith search tool call with the appropriate parameters. For memorysmith search, I should include: - memorysmith:...