Skip to content
STIMSMITH

Search+LLM-based Testing for ARM Simulators

Paper
First seen 8/4/2026
Last seen 8/12/2026
Evidence 11 chunks

NEIGHBORHOOD

No graph connections found for this entity yet. It may appear in future ingestion runs.

explore full graph →

RELATIONSHIPS

27 connections
SearchSYS introduces → 100% 4e
The paper presents and discusses SearchSYS as its primary contribution.
gem5 evaluates → 100% 4e
The paper evaluates SearchSYS by testing the gem5 simulator.
GPT-3.5-turbo evaluates → 97% 2e
The paper empirically investigates six LLMs including GPT-3.5-turbo for input generation.
TinyLlama evaluates → 97% 2e
The paper empirically investigates six LLMs including TinyLlama for input generation.
Phi2 evaluates → 97% 2e
The paper empirically investigates six LLMs including Phi2 for input generation.
Llama2 evaluates → 97% 2e
The paper empirically investigates six LLMs including Llama2 for input generation.
Magicoder evaluates → 97% 2e
The paper empirically investigates six LLMs including Magicoder for input generation.
CodeBooga evaluates → 97% 2e
The paper empirically investigates six LLMs including CodeBooga for input generation.
ARM instruction set architecture ← compares with 95% 2e
The paper compares bug-finding effectiveness between the ARM ISA and X86 ISA.
X86 Instruction Set Architecture ← compares with 95% 2e
The paper compares bug-finding on ARM ISA with prior results on X86 ISA in gem5.
Fuzzing uses → 100% 2e
The paper uses fuzzing as one of the core testing techniques.
TinyLlama uses → 100% 2e
The paper uses TinyLlama as one of the LLMs for seed generation evaluation.
LLM-based Test Generation uses → 100% 2e
The paper uses LLMs for initial seed C code generation.
differential testing uses → 100% 2e
The paper uses differential testing to identify mismatches between simulator and real hardware.
RISC-V mentions → 85% 1e
The paper mentions RISC-V as a potential future ISA target for SearchSYS.
Magicoder uses → 100% 1e
The paper uses Magicoder as one of the LLMs for seed generation evaluation.
Phi uses → 100% 1e
The paper uses Phi as one of the LLMs for seed generation evaluation.
GPT-3.5 uses → 100% 1e
The paper uses GPT-3.5 as one of the LLMs for seed generation evaluation.
Karine Even-Mendoza authored by → 99% 1e
Karine Even-Mendoza is listed as first author of the paper.
Héctor D. Menéndez authored by → 99% 1e
Héctor D. Menéndez is listed as second author of the paper.
W.B. Langdon authored by → 99% 1e
W.B. Langdon is listed as third author of the paper.
Aidan Dakhama authored by → 99% 1e
Aidan Dakhama is listed as fourth author of the paper.
Justyna Petke authored by → 99% 1e
Justyna Petke is listed as fifth author of the paper.
Bobby R. Bruce authored by → 99% 1e
Bobby R. Bruce is listed as sixth author of the paper.
King's College London authored by → 98% 1e
Multiple authors are affiliated with King's College London.
University College London authored by → 98% 1e
Multiple authors are affiliated with University College London.
University of California, Davis authored by → 98% 1e
Bobby R. Bruce is affiliated with UC Davis.