The first SoK on LLM-based AutoPT frameworks provides a six-dimension taxonomy of agent designs and a unified empirical benchmark evaluating 15 frameworks via over 10 billion tokens and 1,500 manually reviewed logs.
hub Canonical reference
Language models are few-shot learners
Canonical reference. 92% of citing Pith papers cite this work as background.
44
Pith papers citing it
Background
92% of classified citations
hub tools
citation-role summary
background 13
citation-polarity summary
roles
background 13representative citing papers
ToBAC is the first backdoor attack on unified autoregressive models, using data or model poisoning to make triggers elicit cross-modal malicious behavior in text and image generation.
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
Evolving-RL jointly optimizes experience extraction and utilization in LLM agents via RL with separate evaluation signals, delivering up to 98.7% relative gains on out-of-distribution tasks in ALFWorld and Mind2Web.
BadDLM implants effective backdoors in diffusion language models across concept, attribute, alignment, and payload targets by exploiting denoising dynamics while preserving clean performance.
MemCompiler reframes memory use as state-conditioned compilation, delivering relevant guidance via text and latent channels to improve embodied agent performance up to 129% and cut latency 60% versus static injection.
LLM surrogate beliefs under sparse observations depend on prompts and query protocols, with structural prompts as priors, pointwise vs joint querying producing different beliefs, and sequential evidence causing non-monotonic updates that affect acquisition and regret.
A multi-agent framework reconstructs the evolutionary graph of post-training LLM datasets, revealing domain patterns like vertical refinement in math data and systemic issues like redundancy and benchmark contamination, then applies it to create a more diverse lineage-aware dataset.
CortexMAE adapts Vision Transformers to fMRI via cortical flat maps, shows power-law scaling on 2.1K hours of data, and outperforms priors on cognitive state decoding while failing to beat a simple functional connectivity baseline on subject-level trait prediction.
Pre-trained LLMs learn to predict HMM-generated sequences via in-context learning, approaching theoretical optimum on synthetic HMMs and matching expert models on real animal decision data.
VISE is the first benchmark for sycophancy in Video-LLMs, with two training-free mitigation strategies based on key-frame selection and internal representation steering.
Orak is a foundational benchmark providing training data, interfaces, and evaluation tools for LLM agents across diverse video game genres.
Chain-of-thought monitoring detects reward hacking in frontier reasoning models, but strong optimization against the monitor produces obfuscated misbehavior that remains hard to detect.
KV cache compression causes task-dependent degradation in high-density reasoning due to disrupted CoT links; ShotKV mitigates this by preserving few-shot examples as indivisible semantic units through phase separation, delivering 9-18% accuracy gains and 11% latency reduction.
TS-Reasoner is a domain-oriented agent using LLMs, computational tools, and error feedback for multi-step time series inference, showing better performance than general LLMs on understanding and reasoning benchmarks.
MuirBench is a new benchmark showing that top multimodal LLMs struggle with robust multi-image understanding, with GPT-4o at 68% and open-source models below 33% accuracy.
Ring Attention uses blockwise computation and ring communication to let Transformers process sequences up to device-count times longer than prior memory-efficient methods.
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
FlashAttention reduces GPU high-bandwidth memory accesses in self-attention via tiling, delivering exact attention with lower IO complexity, 2-3x wall-clock speedups on models like GPT-2, and the ability to train on sequences up to 64K long.
A transformer-based in-context learning model predicts continental-scale subsurface temperatures from sparse borehole observations, outperforming physics and interpolation baselines while adapting to new regions with 20 examples.
Invaria trains point cloud encoders with next-resolution prediction to learn scale and density invariant features, yielding higher mIoU on ScanNet under lower resolution and scaled objects while using a smaller model.
DarkLLM trains an LLM to generate language-driven adversarial perturbations that unify targeted, untargeted, segmentation, and multi-model attacks on foundation models.
CASCADE enables LLMs to continually adapt at deployment via case-based episodic memory and contextual bandits, improving macro-averaged success by 20.9% over zero-shot on 16 tasks spanning medicine, law, code, and robotics.
Intern-Atlas constructs a methodological evolution graph with 9.4 million edges from 1.03 million AI papers to capture how methods emerge, adapt, and transition, enabling better idea evaluation and generation for AI-driven research.
citing papers explorer
-
State-Centric Decision Process
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
-
Evolving-RL: End-to-End Optimization of Experience-Driven Self-Evolving Capability within Agents
Evolving-RL jointly optimizes experience extraction and utilization in LLM agents via RL with separate evaluation signals, delivering up to 98.7% relative gains on out-of-distribution tasks in ALFWorld and Mind2Web.
-
Tracing the Roots: A Multi-Agent Framework for Uncovering Data Lineage in Post-Training LLMs
A multi-agent framework reconstructs the evolutionary graph of post-training LLM datasets, revealing domain patterns like vertical refinement in math data and systemic issues like redundancy and benchmark contamination, then applies it to create a more diverse lineage-aware dataset.
-
Orak: A Foundational Benchmark for Training and Evaluating LLM Agents on Diverse Video Games
Orak is a foundational benchmark providing training data, interfaces, and evaluation tools for LLM agents across diverse video game genres.
-
Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation
Chain-of-thought monitoring detects reward hacking in frontier reasoning models, but strong optimization against the monitor produces obfuscated misbehavior that remains hard to detect.
-
CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment
CASCADE enables LLMs to continually adapt at deployment via case-based episodic memory and contextual bandits, improving macro-averaged success by 20.9% over zero-shot on 16 tasks spanning medicine, law, code, and robotics.
-
Intern-Atlas: A Methodological Evolution Graph as Research Infrastructure for AI Scientists
Intern-Atlas constructs a methodological evolution graph with 9.4 million edges from 1.03 million AI papers to capture how methods emerge, adapt, and transition, enabling better idea evaluation and generation for AI-driven research.
-
SGLang: Efficient Execution of Structured Language Model Programs
SGLang is a new system that speeds up structured LLM programs by up to 6.4x using RadixAttention for KV cache reuse and compressed finite state machines for output decoding.
-
CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society
CAMEL proposes a role-playing framework with inception prompting that enables autonomous multi-agent cooperation among LLMs and generates conversational data for studying their behaviors.
-
EGL-SCA: Structural Credit Assignment for Co-Evolving Instructions and Tools in Graph Reasoning Agents
EGL-SCA co-evolves instructions and tools via structural credit assignment in graph reasoning agents and reports 92% average success on four benchmarks.
-
GENIUS: An Agentic AI Framework for Autonomous Design and Execution of Simulation Protocols
GENIUS is an agentic AI framework that automates generation, validation, and repair of Quantum ESPRESSO DFT input files, succeeding on ~80% of 295 benchmarks with 76% autonomous repairs and lower cost than LLM-only baselines.
-
Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignment
Survey organizes LLM trustworthiness into seven categories and 29 sub-categories, measures eight sub-categories on popular models, and finds that more aligned models generally score higher but with varying effectiveness.