Title it a model card and drop the research-artifact tag, now that a checkpoint ships
Browse files
README.md
CHANGED
|
@@ -9,7 +9,7 @@ inference: false
|
|
| 9 |
datasets:
|
| 10 |
- AgentsSci/EMNLP_Cost-Aware-Protocol-Routing
|
| 11 |
tags:
|
| 12 |
-
-
|
| 13 |
- llm-routing
|
| 14 |
- multi-agent-systems
|
| 15 |
- cost-aware-inference
|
|
@@ -23,7 +23,7 @@ language:
|
|
| 23 |
- en
|
| 24 |
---
|
| 25 |
|
| 26 |
-
# Cost-Aware Protocol Routing —
|
| 27 |
|
| 28 |
**EMNLP 2026 Accepted Paper**
|
| 29 |
*LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks*
|
|
@@ -36,7 +36,7 @@ language:
|
|
| 36 |
|
| 37 |
---
|
| 38 |
|
| 39 |
-
## Read this first: what this
|
| 40 |
|
| 41 |
**This is not a language model.** It contains no foundation-model weights. The
|
| 42 |
paper evaluates two third-party solver families — `openai/gpt-oss-120b` and
|
|
|
|
| 9 |
datasets:
|
| 10 |
- AgentsSci/EMNLP_Cost-Aware-Protocol-Routing
|
| 11 |
tags:
|
| 12 |
+
- protocol-routing
|
| 13 |
- llm-routing
|
| 14 |
- multi-agent-systems
|
| 15 |
- cost-aware-inference
|
|
|
|
| 23 |
- en
|
| 24 |
---
|
| 25 |
|
| 26 |
+
# Cost-Aware Protocol Routing — router model card
|
| 27 |
|
| 28 |
**EMNLP 2026 Accepted Paper**
|
| 29 |
*LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks*
|
|
|
|
| 36 |
|
| 37 |
---
|
| 38 |
|
| 39 |
+
## Read this first: what this model is, and is not
|
| 40 |
|
| 41 |
**This is not a language model.** It contains no foundation-model weights. The
|
| 42 |
paper evaluates two third-party solver families — `openai/gpt-oss-120b` and
|