Upload 6 files

README.md

@@ -1,3 +1,115 @@
 ---
 license: mit
+language:
+  - en
+pretty_name: Serving-Stack Confounds in Local Tool-Use Evaluation
+size_categories:
+  - n<1K
+task_categories:
+  - text-generation
+tags:
+  - tool-calling
+  - function-calling
+  - llm-evaluation
+  - llm-agents
+  - ollama
+  - vllm
+  - sglang
+  - llama-cpp
+configs:
+  - config_name: serving_modes
+    default: true
+    data_files: serving_modes.csv
+  - config_name: chain_compare
+    data_files: chain_compare.csv
+  - config_name: crossstack
+    data_files: crossstack.csv
+  - config_name: humaneval_probe
+    data_files: humaneval_probe.csv
 ---
+
+# Serving-Stack Confounds in Local Tool-Use Evaluation
+
+Result tables for **Measuring the Serving Stack Instead of the Model: Hidden Confounds in
+Local Tool-Use Evaluation** (Lijuan Tang, Yuemeng Zheng; REALM Workshop @ EMNLP 2026).
+
+- Paper: [ACL Anthology](https://aclanthology.org/2026.realm-1.7/) · [arXiv:2609.26693](https://arxiv.org/abs/2609.26693)
+- Code, configs, and reproduction steps: [github.com/LijuanTang94/serving-confound-repo](https://github.com/LijuanTang94/serving-confound-repo)
+
+![Per-seed protocol fidelity under three serving conditions](serving_modes.png)
+
+## What this measures
+
+**Protocol fidelity** = the rate of valid, in-schema tool calls over the turns the model
+actually produced. Requests the server refused (for example Ollama's HTTP 400
+`does not support tools` for Phi-3 and Gemma-3) are counted separately as rejected and
+excluded from the denominator. A naive analysis that counts them as model non-calls
+reports a capable model at 0% even though it never ran.
+
+Serving conditions:
+
+| condition | `tools=` sent | text tool list in prompt |
+|---|---|---|
+| `native` (default) | yes | no |
+| `native_hint` | yes | yes |
+| `text_tools` / `tt` | no | yes |
+
+All local models run on Ollama 0.30.8 with default (Q4_K_M) quantization;
+`deepseek-v4-flash` is a cloud anchor. A "seed" indexes a task instance; decoding is
+sampled at T=1.0.
+
+## Configs
+
+### `serving_modes` (default): aggregation task, 8 seeds, three conditions
+
+For each condition prefix (`native`, `native_hint`, `text_tools`):
+
+| column | meaning |
+|---|---|
+| `<cond>_seed` | per-seed mean fidelity |
+| `<cond>_lo`, `<cond>_hi` | seed-level bootstrap 95% CI |
+| `<cond>_pooled` | turn-pooled fidelity |
+| `<cond>_n` | model-produced turns in the pool |
+| `<cond>_rej` | rejected requests (excluded from the denominator) |
+| `<cond>_n_seeds` | seeds with at least one produced turn |
+
+Empty cells mean the condition was not run or the rate is undefined (every request was
+rejected).
+
+### `chain_compare`: dependency-chain task, 4 seeds
+
+`native_*` and `tt_*` (text-tools) with `_seed` (per-seed mean), `_pool` (turn-pooled),
+`_n` (turns), `_rej` (rejected requests).
+
+### `crossstack`: same `tools=` request on four serving stacks
+
+One row per model; each cell is the outcome on that stack: `200_native` (native
+`tool_calls`), `200_text` / `200_text_json_block` (call returned as text),
+`200_runs_text` (request accepted, model runs), `200_prose_nocall` / `200_text_prose`
+(prose, no call), `HTTP400_rejected` (Ollama per-model gating), `HTTP400_no_parser`
+(vLLM without `--enable-auto-tool-choice --tool-call-parser`), `not_probed`, `n/a`.
+`vllm_configured` is vLLM with tool parsing enabled. The vLLM and SGLang probes cover
+only the models shown.
+
+### `humaneval_probe`: single-turn replication on 6 HumanEval problems
+
+`native_gating` is the Ollama response to `tools=` (`-` = not probed), then
+`text_tools_valid` of `n` problems produced a valid call under the text-tools protocol.
+
+## Citation
+
+```bibtex
+@inproceedings{tang-zheng-2026-measuring,
+    title = "Measuring the Serving Stack Instead of the Model: Hidden Confounds in Local Tool-Use Evaluation",
+    author = "Tang, Lijuan  and
+      Zheng, Yuemeng",
+    booktitle = "Proceedings of the Second Workshop for {RE}search on Agent Language Models ({REALM} 2026)",
+    month = oct,
+    year = "2026",
+    address = "Budapest, Hungary",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/2026.realm-1.7/",
+    doi = "10.18653/2026.realm-1.7",
+    pages = "86--94"
+}
+```

chain_compare.csv

@@ -0,0 +1,10 @@
+model,native_seed,native_pool,native_n,native_rej,tt_seed,tt_pool,tt_n,tt_rej
+qwen2.5-coder:0.5b,0.0,0.0,542,0,0.438,0.8,15,1
+qwen2.5-coder:1.5b,0.502,0.506,83,1,0.812,0.875,16,2
+qwen2.5-coder:3b,,,,,0.867,0.833,12,0
+qwen2.5-coder:7b,,,,,0.847,0.885,26,1
+qwen2.5-coder:14b,,,,,1.0,1.0,29,3
+llama3.2:latest,,,,,0.125,0.2,5,0
+phi3:latest,,,,,0.5,0.5,4,0
+gemma3:4b,,,,,1.0,1.0,4,4
+gemma3:270m,,,,,0.0,0.0,4,0

crossstack.csv

@@ -0,0 +1,6 @@
+model,ollama_tools,llamacpp_tools,vllm_default,vllm_configured,sglang_default
+qwen2.5-coder:0.5b,200_text,200_text,HTTP400_no_parser,200_text_json_block,200_text_json_block
+llama3.2,200_native,200_native,HTTP400_no_parser,not_probed,not_probed
+phi3,HTTP400_rejected,200_runs_text,HTTP400_no_parser,200_prose_nocall,200_text_prose
+gemma3:270m,HTTP400_rejected,200_runs_text,not_probed,not_probed,not_probed
+gemma2:2b,n/a,n/a,HTTP400_no_parser,not_probed,not_probed

humaneval_probe.csv

@@ -0,0 +1,8 @@
+model,native_gating,text_tools_valid,n,text_tools_rate
+qwen2.5-coder:0.5b,200_text,4,6,0.667
+qwen2.5-coder:1.5b,-,5,6,0.833
+qwen2.5-coder:3b,-,6,6,1.000
+llama3.2:latest,200_text,6,6,1.000
+phi3:latest,HTTP400_rejected,4,6,0.667
+gemma3:4b,-,2,6,0.333
+gemma3:270m,HTTP400_rejected,0,6,0.000

serving_modes.csv

@@ -0,0 +1,11 @@
+model,native_seed,native_lo,native_hi,native_pooled,native_n,native_rej,native_n_seeds,native_hint_seed,native_hint_lo,native_hint_hi,native_hint_pooled,native_hint_n,native_hint_rej,native_hint_n_seeds,text_tools_seed,text_tools_lo,text_tools_hi,text_tools_pooled,text_tools_n,text_tools_rej,text_tools_n_seeds
+qwen2.5-coder:0.5b,0.0,0.0,0.0,0.0,9,0,8,0.346,0.067,0.659,0.287,317,2,8,0.344,0.094,0.594,0.852,54,1,8
+qwen2.5-coder:1.5b,0.375,0.125,0.75,0.237,519,3,8,0.583,0.312,0.833,0.906,53,1,8,0.742,0.492,0.975,0.955,110,1,8
+qwen2.5-coder:3b,0.228,0.0,0.549,0.786,42,0,8,0.823,0.575,0.979,0.96,176,3,8,0.837,0.702,0.958,0.896,77,1,8
+qwen2.5-coder:7b,0.589,0.319,0.837,0.793,174,2,8,0.885,0.781,0.969,0.8,35,0,8,0.912,0.825,0.979,0.889,36,0,8
+qwen2.5-coder:14b,0.595,0.383,0.772,0.737,57,0,8,0.796,0.625,0.944,0.708,89,0,8,0.915,0.856,0.979,0.882,34,0,8
+llama3.2:latest,0.6,0.375,0.8,0.625,16,0,8,0.823,0.709,0.921,0.892,65,0,8,0.438,0.312,0.5,0.467,15,0,8
+phi3:latest,,,,0.0,0,8,0,,,,,,,,0.375,0.125,0.75,0.3,10,0,8
+gemma3:4b,,,,0.0,0,8,0,,,,,,,,1.0,1.0,1.0,1.0,8,8,8
+gemma3:270m,,,,0.0,0,8,0,,,,,,,,0.375,0.125,0.75,0.375,8,4,8
+deepseek-v4-flash,1.0,1.0,1.0,1.0,61,0,8,,,,,,,,1.0,1.0,1.0,1.0,57,0,8