Latest AI updates

ETAPS: Gradual Pruning with Sparse Adaptation Achieves High Quality at Extreme Sparsity

Researchers propose ETAPS, a pruning method that gradually tightens a global sparsity constraint while training a sparse delta with a small fixed budget. This approach avoids model-sized optimizer states, enabling efficient pruning of LLMs at extreme sparsity levels. Across models from 125M to 13B parameters, ETAPS achieves lower perplexity than the prior state-of-the-art ELSA at sparsities from 50% to 90%, with up to 42% perplexity reduction. It also reduces cost significantly, pruning LLaMA2-7B on a single H200 with 5x smaller persistent state and 49% fewer GPU-hours compared to ELSA.

Why it matters: This paper presents a concrete, recent advancement in LLM pruning that improves both quality and efficiency, offering practical value for deploying large models.

GOAL Targets Harmless-Query Refusals Through Gradient and Representation Changes

A new paper examines why safety-aligned language models reject harmless prompts, reporting increasing overlap between harmful and over-refusal representations in deeper layers and gradient interactions that favor refusal. The researchers propose GOAL, combining gradient orthogonalization with contrastive learning to address these effects. They report reduced over-refusal across multiple models and benchmarks while preserving safety performance and general capabilities, though the supplied abstract provides no quantitative results.

Why it matters: The work offers a concrete training approach to reducing unnecessary refusals without sacrificing safety, addressing a practical limitation of aligned language models.

HEMO Uses LLMs to Evolve the Search Process for Multi-Objective Optimization

Researchers introduce HEMO, a framework that adapts the harness used in LLM-based automatic operator design for multi-objective combinatorial optimization. It updates modules governing experience retrieval, operator improvement, parent selection and search orchestration, guided by solution quality and runtime while preserving feedback in persistent memory. Experiments across eight tasks and two evolutionary backbones report strong performance against conventional evolutionary, neural and LLM-based baselines, along with transfer to related tasks.

Why it matters: HEMO offers a concrete approach to improving the orchestration of LLM-driven optimization, with evaluation across multiple tasks and search backbones.

Mistral Large 4 'Le Chonk' Released on Hugging Face

Mistral AI has released Mistral Large 4, nicknamed 'Le Chonk', on Hugging Face. The model is available for download and integration through the platform. This release marks a significant addition to the open-source AI model ecosystem.

Why it matters: Provides direct access to a new major model release, valuable for developers and researchers seeking cutting-edge open-weight AI.

SYNAPSE: Neuro-Symbolic Visual Thought-to-Text Decoding via Topological Semantic Denoising

Recent advancements in Large Language Models (LLMs) have accelerated research into open-vocabulary electroencephalography (EEG)-to-text decoding, a task that translates non-invasive neural activity recorded during visual perception into coherent natu<!--autointro-->

China&#039;s Top 5 AI Large Language Models: A Comprehensive Comparison

The competition among large language models (LLMs) in China has entered a phase where each model excels in specific areas. Rather than asking &quot;which one is the best?&quot;, the more practical question is &quot;which one is best for&nbsp;your&nbs<!--autointro-->

SPEX and ProxySPEX: Scalable Interaction Discovery for LLM Interpretability

Berkeley AI Research (BAIR) presents SPEX and ProxySPEX, algorithms that identify influential interactions among features, training data points, and model components at scales far beyond prior methods. SPEX exploits sparsity and low-degreeness to reframe interaction search as a sparse recovery problem; ProxySPEX adds hierarchy to achieve similar performance with roughly 10x fewer ablations. The methods are demonstrated on feature attribution (long-context sentiment, a modified trolley problem),

Why it matters: It offers a scalable, open-source interpretability framework with concrete demonstrations across feature, data, and component attribution, useful for researchers and practitioners working on model transparency.