YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Reason-First Program
Concept-Guided Program Space Exploration
When an LLM fills a stub function, there exists a possible program space of valid implementations. Any implementation that satisfies the spec, the formal constraints, and the observable effects is "aligned" — its exact shape shouldn't matter. What does matter is the set of concepts that characterize distinct regions of that space.
This framework discovers those concepts, projects programs into a concept-guided embedding space, and provides a query language for steering generation.
Architecture
┌─────────────────────────────────────────────────────────────────┐
│ @reason_first(spec="...") │
│ def process(items): ... │
│ ┌──────────┐ │
│ │ STUB │ │
│ └────┬─────┘ │
│ │ │
│ ┌───────────────┼───────────────┐ │
│ ▼ ▼ ▼ │
│ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │
│ │ Model A │ │ Model B │ │ Model C │ Stage 1 │
│ │ T=0.2..1.2 │ │ T=0.2..1.2 │ │ T=0.2..1.2 │ Sampling │
│ └──────┬───────┘ └──────┬───────┘ └──────┬───────┘ │
│ └────────────────┼────────────────┘ │
│ ▼ │
│ ┌─────────────────────────┐ │
│ │ PROGRAM SPACE │ │
│ │ (valid implementations)│ │
│ │ DA@K diversity metrics │ │
│ └────────────┬────────────┘ │
│ │ │
│ ┌───────────────┼───────────────┐ │
│ ▼ ▼ ▼ │
│ ┌─────────────┐ ┌──────────────┐ ┌─────────────┐ │
│ │ Behavioral │ │ SAE-Based │ │ Abstraction │ Stage 2 │
│ │ (execution) │ │(hidden state)│ │ (AST frag.) │ Discovery │
│ └──────┬──────┘ └──────┬───────┘ └──────┬──────┘ │
│ └───────────────┼────────────────┘ │
│ ▼ │
│ ┌──────────────────────────┐ │
│ │ CONCEPT SET │ │
│ │ {recursive, mutation, │ │
│ │ hash_based, sorting, │ │
│ │ fast_execution, ...} │ │
│ └────────────┬─────────────┘ │
│ │ │
│ ┌────────────┼────────────┐ │
│ ▼ ▼ ▼ │
│ ┌──────────────┐ ┌─────────┐ ┌──────────┐ │
│ │ CB-AE │ │ GCAV │ │ MSRS │ Stage 3 │
│ │ Bottleneck │ │ Vectors │ │ Orthog. │ Embedding │
│ └──────┬───────┘ └────┬────┘ └────┬─────┘ │
│ └──────────────┼───────────┘ │
│ ▼ │
│ ┌──────────────────────────┐ │
│ │ CONCEPT EMBEDDING SPACE │ │
│ │ (interpretable axes, │ │
│ │ verifiable alignment) │ │
│ └────────────┬─────────────┘ │
│ │ │
│ ▼ │
│ ┌──────────────────────────┐ │
│ │ QUERY LANGUAGE │ Stage 4 │
│ │ recursive=0.8, │ Steering │
│ │ mutation=-0.5, │ │
│ │ fast_execution=0.7 │ │
│ └──────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
Quick Start
from reason_first_program import (
ProgramSpace, Program, UnifiedConceptDiscovery, SteeringEngine
)
from reason_first_program.stub import Stub, StubConstraints
from reason_first_program.program_space import execute_program
# 1. Define a stub
stub = Stub(
name="two_sum",
source='def two_sum(nums: list[int], target: int) -> list[int]:\n ...',
signature="(nums: list[int], target: int) -> list[int]",
constraints=StubConstraints(decorator_spec="Find two indices that sum to target"),
test_inputs=[{"nums": [2, 7, 11, 15], "target": 9}],
)
# 2. Build program space (add implementations)
space = ProgramSpace(stub)
for src in implementations:
p = Program(source=src, full_source=src, stub_id=stub.stub_id)
p = execute_program(p, stub, stub.test_inputs)
space.add(p)
# 3. Discover concepts
discovery = UnifiedConceptDiscovery()
concepts = discovery.discover(space)
# 4. Query & steer
engine = SteeringEngine(concepts)
results = engine.select(space, "uses_dict=1.0, uses_mutation=-1.0", top_k=3)
Using the @reason_first Decorator
from reason_first_program import reason_first
@reason_first(
spec="Sort items by priority, breaking ties by recency",
postconditions=["output is sorted", "all input items present in output"],
)
def process_queue(items: list) -> list:
#> stable sort; O(n log n); must preserve Item identity
...
Concept Discovery Methods
1. Behavioral Concepts (AST + Execution)
Discovers concepts from code structure and execution traces:
- Algorithmic patterns:
uses_recursion,uses_iteration,uses_list_comprehension - Data structure usage:
uses_dict,uses_set,uses_heap - Control flow:
uses_early_return,single_return,uses_generator - Mutation pattern:
uses_mutation(in-place) vs immutable - Performance:
fast_execution(below-median execution time)
2. SAE-Based Concepts (Hidden States)
Trains a Sparse Autoencoder on code LLM hidden states (requires GPU):
- Based on CB-SAE and DN-CBM
- Auto-names neurons against a code concept vocabulary
- Discovers implicit representational concepts the LLM uses internally
3. Structural Abstraction Concepts (AST Fragments)
Finds common program fragments across implementations:
Diversity Metrics
report = space.diversity_report()
# {
# "total_programs": 10,
# "valid_programs": 10,
# "functionally_unique_clusters": 3,
# "da_at_5": 2.8, # Expected distinct algorithms in 5 samples
# "entropy_diversity": 2.1, # Entropy-based diversity index
# }
From AlgoDiv (2503.00691): DA@K = Σ_m (1 - C(N-s_m, K) / C(N, K))
Embedding & Steering
GCAV Concept Vectors
Based on GCAV (2501.05764): e' = e + ε · v_concept
from reason_first_program.embeddings import GCAVEmbedding
gcav = GCAVEmbedding()
gcav.train_all(concepts, features, programs)
steered = gcav.multi_steer(features, {
"uses_recursion": 0.8, "fast_execution": 0.6, "uses_mutation": -0.3
})
MSRS Orthogonal Steering
Based on MSRS (2508.10599): orthogonal subspaces prevent concept interference.
Query Language
engine = SteeringEngine(concepts)
# Weight-based queries
results = engine.select(space, "uses_recursion=0.8, uses_mutation=-0.5")
# Constraint queries
results = engine.select(space, "uses_dict > 0.5 AND fast_execution > 0.7")
# Concept negation
results = engine.select(space, "NOT uses_mutation")
# Programmatic
query = engine.query_language.build(uses_recursion=0.8, fast_execution=0.6)
results = engine.select(space, query)
Concept Lattice (FCA)
Builds a Formal Concept Analysis lattice:
- Extent: set of programs sharing certain properties
- Intent: set of concepts shared by those programs
lattice = concepts.concept_lattice()
for extent, intent in lattice[:5]:
print(f"Concepts {sorted(intent)} → {len(extent)} programs")
Theoretical Foundations
| Method | Paper | Key Contribution |
|---|---|---|
| Program space diversity | AlgoDiv | DA@K metric, AlgoSim clustering |
| SFS scattering | SFS | Diverse algorithmic direction discovery |
| SAE concept discovery | CB-SAE | Concept Bottleneck Sparse Autoencoders |
| Concept naming | DN-CBM | Discover-then-Name concept pipeline |
| Concept vectors | GCAV | Concept Activation Vectors for steering |
| Multi-concept steering | MSRS | Orthogonal subspace steering |
| Execution semantics | TRACED | Execution-trace-aware representations |
| Library learning | DreamCoder / LILO | Abstraction discovery from program corpora |
| Latent program space | LPN | Continuous latent program representations |
| Typed holes | Hazel | Spec-constrained completion spaces |
Running the Experiment
pip install -e .
python -m reason_first_program.experiment
Demonstrates the full pipeline on hand-crafted two_sum and flatten implementations:
- Program space construction & diversity metrics
- Behavioral + structural concept discovery (31 concepts for two_sum, 30 for flatten)
- Concept-guided embedding & alignment verification
- Query language selection & concept boundary exploration
- Formal concept lattice construction
Installation
# Core (concept discovery + steering, no GPU needed)
pip install -e .
# With LLM sampling support
pip install -e ".[llm]"
# With SAE concept discovery (requires GPU)
pip install -e ".[sae]"
# Everything
pip install -e ".[all]"
License
Apache 2.0