Yixin Cao, Changyi Xiao, Caijun Xu
We lifted 13 functions out of this paper's own repositories and ran 6 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| changyi7231/CER | canonical | 6 of 13 |
| Function | Status | Where it lives |
|---|---|---|
| exact_match | Ran | changyi7231/CER/recipe/cer/src/reward_manager.py code served (permissive licence) · get_code("bacda3216fdce102") |
| extract_boxed_answer | Ran | changyi7231/CER/recipe/cer/src/data_preparation.py code served (permissive licence) · get_code("904946a1f147981f") |
| extract_solution_answer | Ran | changyi7231/CER/recipe/cer/src/cer_ray_trainer.py code served (permissive licence) · get_code("5285b1569c3d6597") |
| get_prompt | Ran | changyi7231/CER/recipe/cer/src/data_preparation.py code served (permissive licence) · get_code("a6f637988f82514b") |
| is_ipv4 | Ran | changyi7231/CER/verl/utils/net_utils.py code served (permissive licence) · get_code("735e93af30b15fa9") |
| unpad_dataproto | Ran | changyi7231/CER/verl/protocol.py code served (permissive licence) · get_code("25f0ea3f460f6ce1") |
| get_generation_config | Not yet run | changyi7231/CER/verl/utils/model.py code served (permissive licence) · get_code("c9d58036da0286f2") |
| get_huggingface_actor_config | Not yet run | changyi7231/CER/verl/utils/model.py code served (permissive licence) · get_code("59904a8955b34252") |
| get_weight_loader | Not yet run | changyi7231/CER/verl/models/weight_loader_registry.py code served (permissive licence) · get_code("acf240079d434746") |
| get_weight_saver | Not yet run | changyi7231/CER/verl/models/weight_loader_registry.py code served (permissive licence) · get_code("55f217ce71432b58") |
| is_ipv6 | Not yet run | changyi7231/CER/verl/utils/net_utils.py code served (permissive licence) · get_code("705afe175ad8e404") |
| load_state_dict_to_megatron_gptmodel | Not yet run | changyi7231/CER/verl/models/mcore/loader.py code served (permissive licence) · get_code("1853101234cf952e") |
| squeeze | Not yet run | changyi7231/CER/verl/utils/model.py code served (permissive licence) · get_code("3b15e2ac7497c441") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective in enhancing the reasoning capabilities of large language models, particularly in domains such as mathematics where reliable rule-based verifiers can be constructed. However, the reliance on handcrafted, domainspecific verification rules substantially limits the applicability of RLVR to general reasoning domains with free-form answers, where valid answers often exhibit significant variability, making it difficult to establish complete and accurate rules. To address this limitation, we propose Conditional Expectation Reward (CER), which leverages the large language model itself as an implicit verifier, and is therefore applicable to general domains and eliminates the need for external verifiers or auxiliary models. CER is defined as the expected likelihood of generating the reference answer conditioned on the generated answer. In contrast to rule-based verifiers that yield binary feedback, CER provides a soft, graded reward signal that reflects varying degrees of correctness, making it better suited to tasks where answers vary in correctness. Experimental results demonstrate that CER is effective across a wide range of reasoning tasks, spanning both mathematical and general domains, indicating that CER serves as a flexible and general verification mechanism. The code is available at https: //github.com/changyi7231/CER.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2603.10624")
get_code_for_paper("2603.10624")
have("2603.10624")
Connect an agent — have() is free.