IX2-0339
① SA Source
- Source: 開啟完整 SA 文章
- Section:
Accuracy Evaluations - Line hint:
560
Context Before
Source: SemiAnalysis InferenceX ↗
Accuracy Evaluations
Evidence
For example, this additional layer of checks has helped us discover issues with some DP attention implementation for GPT-OSS
Context After
Each representative throughput config now has an associated numerical accuracy check. Currently we are only using GSM8k, but being a very easy benchmark, the evaluation scores may not change much from differences in numerical calculation, and a harder benchmark may have a larger delta with respect to numerical accuracy. Thus, we plan to expand towards harder ones in the future, such as GPQA, HLE, MATH-500, SWE-Bench verified.
Another form of performance-accuracy tradeoff is quantization. Serving models at lower precision may result in worse model outputs. For DeepSeek R1, FP8 runs have very slightly higher evaluation scores than FP4. Note that GSM8k evals are saturated and often during QAT/PAT it is calibrated to common popular GSM8k, MATH-500, etc, leading to sometimes evals showing great results while real world end user evaluation being subpar. If we want to be part of the team to figure out how to properly evaluate inference engine accuracy, apply to join the mission here ↗.
② Atomic Claim
例如加入這層 accuracy checks 後，InferenceX 曾發現 GPT-OSS 某些 DP attention implementation 的問題。
- Epistemic Mode:
ASSERTED - Mapping Status:
PARTIAL
③ Semantic Frame
{
"attribute": "UNSPECIFIED_ATTRIBUTE",
"context_nodes": [],
"entity": {
"id": "04_knowledge_base/GPT-OSS",
"label": "GPT-OSS"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "例如加入這層 accuracy checks 後，InferenceX 曾發現 GPT-OSS 某些 DP attention implementation 的問題。"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | GPT-OSS | GPT-OSS |
⑤ Human Review
請在 Properties 逐項確認：
- 原文 → Atomic Claim 是否忠實
- Atomic Claim → Semantic Frame 是否忠實
- Canonical Entity mapping 是否正確
- Epistemic mode 是否保留原文語氣
- 最後選擇
review_action
Review state
Markdown 內文不是正式 approval。只有 Apply bridge 寫入的 Decision Ledger event 才是正式決策。