Difference between revisions of "Increasing AI Intelligence"

From GISAXS
Jump to: navigation, search
(→‎Usage of Reasoning Compute)
(→‎Proactive Search)
 
(2 intermediate revisions by the same user not shown)
Line 31: Line 31:
 
=Proactive Search=
 
=Proactive Search=
 
Compute expended after training, but before inference.
 
Compute expended after training, but before inference.
 +
 +
===Reinforcement Learning===
 +
* 2025-04: DeepSeek: [https://arxiv.org/abs/2504.02495 Inference-Time Scaling for Generalist Reward Modeling]
  
 
===Training Data (Data Refinement, Synthetic Data)===
 
===Training Data (Data Refinement, Synthetic Data)===
Line 195: Line 198:
 
* 2025-03: [https://arxiv.org/abs/2503.17352 OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement] ([https://github.com/yihedeng9/OpenVLThinker code])
 
* 2025-03: [https://arxiv.org/abs/2503.17352 OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement] ([https://github.com/yihedeng9/OpenVLThinker code])
 
* 2025-03: [https://arxiv.org/abs/2503.19877 Scaling Evaluation-time Compute with Reasoning Models as Process Evaluators]
 
* 2025-03: [https://arxiv.org/abs/2503.19877 Scaling Evaluation-time Compute with Reasoning Models as Process Evaluators]
 +
* 2025-03: [https://arxiv.org/abs/2503.23513 RARE: Retrieval-Augmented Reasoning Modeling]
  
 
===Model Merging===
 
===Model Merging===
Line 215: Line 219:
 
* 2025-02: [https://www.arxiv.org/abs/2502.08606 Distillation Scaling Laws]
 
* 2025-02: [https://www.arxiv.org/abs/2502.08606 Distillation Scaling Laws]
 
* 2025-03: [https://arxiv.org/abs/2503.10061 Compute Optimal Scaling of Skills: Knowledge vs Reasoning]
 
* 2025-03: [https://arxiv.org/abs/2503.10061 Compute Optimal Scaling of Skills: Knowledge vs Reasoning]
 +
* 2025-03: [https://arxiv.org/abs/2504.00294 Inference-Time Scaling for Complex Tasks: Where We Stand and What Lies Ahead]
  
 
====(Optimal) Usage of Reasoning Compute====
 
====(Optimal) Usage of Reasoning Compute====

Latest revision as of 10:07, 4 April 2025

Reviews

World Model

Prompt Engineering

Thought Templates

Automatic Prompt Optimization

Fine Tuning

Proactive Search

Compute expended after training, but before inference.

Reinforcement Learning

Training Data (Data Refinement, Synthetic Data)

Re-captioning

Generate consistent plans/thoughts

  • 2024-08: Mutual Reasoning Makes Smaller LLMs Stronger Problem-Solvers (code)
    • (Microsoft) rStar is a self-play mutual reasoning approach. A small model adds to MCTS using some defined reasoning heuristics. Mutually consistent trajectories can be emphasized.
  • 2024-09: Self-Harmonized Chain of Thought
    • Produce refined chain-of-thought style solutions/prompts for diverse problems. Given a large set of problems/questions, first aggregated semantically, then apply zero-shot chain-of-thought to each problem. Then cross-pollinate between proposed solutions to similar problems, looking for refined and generalize solutions.
  • 2024-11: LLMs Do Not Think Step-by-step In Implicit Reasoning
    • They argue that models trained to reproduce CoT outputs do not, internally, perform stepwise reasoning (with intermediate representations); this suggests that explicit CoT could be superior to implicit CoT.

Sampling

Automated prompt generation

Distill inference-time-compute into model

CoT reasoning model

See also: AI tools > LLM > Open-weights LLM > Reasoning

Scaling

Inference Time Compute

Methods

Review

In context learning (ICL), search, and other inference-time methods

Inference-time Sampling

Inference-time Gradient

Self-prompting

Retrieval or Memory

In-context thought

Naive multi-LLM (verification, self-critique, majority voting, best-of-N, etc.)

Multi-LLM (multiple comparisons, branching, etc.)

Iteration (e.g. neural-like layered blocks)

Iterative reasoning via graphs

Monte Carlo Tree Search (MCTS)

Other Search

Chain-of-Thought Reasoning

Model Merging

Meta-methods

Analysis

Scaling

(Optimal) Usage of Reasoning Compute

Usage of Training Data

  • 2025-02: LIMO: Less is More for Reasoning (surprisingly easy generalization, from very few reasoning training examples; model can go from knowledge-retrieval to diverse reasoning using curated examples)

Theory

Expending compute works

Compute.png

Pragmatics

Code for Inference-time Compute

  • optillm: Inference proxy which implements state-of-the-art techniques to improve accuracy and performance of LLMs (improve reasoning over coding, logical and mathematical queries)

Interact with Environment

Memory

Tool Use

Integrated

Multi-agent Effort (and Emergent Intelligence)

ML-like Optimization of LLM Setup

Limitations/Requirements

Creativity

See: AI creativity

See Also

Retrieved from "http://gisaxs.com/index.php?title=Increasing_AI_Intelligence&oldid=7451"