Volver a Noticias
InnovaciónAI Understanding sesión informativa

Los investigadores presentan Env-Rethink para mejorar el rendimiento de los agentes de IA en entornos complejos

A new system called Env-Rethink uses a 27B parameter model to help AI agents navigate fragmented, noisy, and evolving digital environments, improving task success rates by over 15%.

4 min readRead the primary source
Source-provided image accompanying Researchers introduce Env-Rethink to improve AI agent performance in complex environments
Documento de fuente primariaFuente registrada
Editor
arxiv.org
Enlace fuente
arxiv.orghttps://arxiv.org/abs/2609.29773
Tipo de fuente
Documento principal: un anuncio oficial, documento, archivo o página propia que leemos directamente.
ContextoEntiende esto en 60 segundos

Empieza aquí

Términos clave

Agente de IA
Un sistema de software que puede observar, razonar y tomar acciones para lograr un objetivo, a menudo utilizando herramientas y memoria.
Parámetro
Un peso aprendido dentro de un modelo que influye en sus resultados.
Latencia
El tiempo entre el envío de una solicitud y la recepción del resultado del modelo.
Ponte a pruebaPrueba de agentes de IA

que paso

Researchers have introduced Env-Rethink, a system designed to address the limitations AI agents face when operating in real-world environments that are not optimized for autonomous interaction. The system utilizes a 27B post-trained model to organize fragmented data and adapt to changing environmental conditions, which the authors claim can improve agent performance by over 15.1% across various tasks.

The Env-Rethink system introduces three core capabilities to assist AI agents. First, it constructs 'Collection Maps' to organize related files and 'Event Logs' to track cross-data relationships, effectively creating a structured context for the agent. Second, it employs offline trajectory learning to identify and filter out noise or misleading information within the environment. Third, it creates virtual event histories that modify environmental states, allowing the agent to train on increasingly difficult scenarios.

The system is powered by a 27B post-trained model. According to the researchers, this architecture allows the agent to adaptively manage environments that are otherwise fragmented or prone to conflicting data versions. The study evaluated the system across 30 tasks and nine different models, reporting a 15.1% improvement in rubric pass rates.

The researchers emphasize that real-world environments, such as office workflows or scientific experimentation platforms, are rarely designed for AI interaction. Env-Rethink acts as an intermediary layer that prepares these environments for agentic tasks.

Detalles de la fuente: arxiv.org ↗

Por qué es importante

Los agentes de IA a menudo tienen dificultades cuando se les asigna la tarea de navegar en entornos digitales complejos del mundo real donde la información está dispersa, es contradictoria o está sujeta a cambios constantes. Al proporcionar un mecanismo para estructurar estos datos y simular escenarios en evolución, Env-Rethink ofrece una vía para que los agentes mantengan una mayor precisión en entornos dinámicos. Esta investigación destaca la necesidad crítica de entornos "preparados para los agentes", lo que sugiere que el cuello de botella para la capacidad de los agentes no es sólo el modelo en sí, sino el entorno en el que opera. La capacidad de generar historiales de eventos virtuales "complicados" para la superación personal representa un paso significativo hacia sistemas más robustos y autónomos capaces de manejar flujos de trabajo del mundo real.

The primary contribution of this work is the shift in focus from improving the agent's internal reasoning to improving the agent's environmental context. By addressing the 'Environment Wall,' the researchers provide a framework for agents to operate in messy, real-world data landscapes.

The capability for recursive self-improvement through the generation of synthetic, challenging event histories is a notable development. By creating 'tricky' environments, the system forces the agent to refine its decision-making processes, potentially leading to more resilient autonomous behavior in unpredictable settings.

Interactive Mechanism

Mecanismo interactivo: cómo funciona realmente

Explore la tecnología subyacente detrás de este desarrollo de forma interactiva.

Thinking Budget (Test-Time Tokens):1,024 tokens
Complex Accuracy79%Math & Code Logic
Latency3.2sTime to first full output
Inference Cost$0.0092Per query estimated
Reasoning StyleStep VerificationInternal chain depth
Active Thinking Trace:
1Deconstruct user problem into formal constraints
2Propose candidate hypotheses & step-by-step calculation
3Self-correction: Backtrack and refute subtle edge cases
4Exhaustive consistency check & final output synthesis
Core takeaway: Test-time compute fundamentally changes AI economics. Instead of only scaling during pre-training, giving reasoning models more tokens at inference time allows them to systematically solve PhD-level STEM problems.
Verificación interactiva del concepto+10 Points
AI Agents Quiz

What most distinguishes an AI agent from a basic chatbot?

Qué ver a continuación

The research team notes that current agent performance can drop significantly—from 83.9% to 57.6%—when environments are not optimized. It remains unknown how Env-Rethink will scale when applied to highly specialized or proprietary enterprise environments beyond the 30 tasks tested in the study. Additionally, the long-term safety implications of allowing agents to 'evolve' their own environments through virtual event histories require further scrutiny, particularly regarding how these synthetic environments align with real-world constraints.

The study does not specify the computational overhead required to maintain these Collection Maps and Event Logs in real-time. Future deployments will need to determine if the performance gains justify the additional and resource consumption.

The researchers have not disclosed the specific datasets used for the 30 tasks, nor have they provided information on the availability or licensing of the Env-Rethink system. It is currently unclear if this will be released as an open-source tool or integrated into existing agent platforms.

Guías y cuestionarios relacionados

Agentes de IAModelos de IA explicadosEntrenamiento de IAFuturo de la IAPon a prueba lo que sabes: prueba un cuestionario gratuito sobre IABusque un término de IA en nuestro glosarioSiga el rastreador de lanzamientos de modelos de IA
¿Encontró esto útil?