← Back to all news

Large language models

Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.

29,398 stories · RSS feed

arXiv AI
Sep 30

One Geometry, Different Outcomes: Readout-Dependent Effects of the Modality Gap in Vision-Language Models

The paper investigates how the modality gap— the separation between image and text representations in contrastive vision‑language models—affects different downstream tasks. By showing that a single dominant direction accounts for most of the image‑text mean separation, the authors explain why reducing or removing this gap can improve zero‑shot classification, degrade retrieval, or restore performance depending on the task. The study provides a geometric framework that clarifies when and why gap interventions should be applied in vision‑language systems.

By Aditya Sharma, Divya Saxena
llmsragmultimodalsafety
More like this →
arXiv AI
Sep 30

Render Before Reading: Visual Rendering as a Prompt Injection Defense

The paper investigates how multimodal large language models are more susceptible to prompt injection when adversarial instructions are presented as text rather than as non-textual inputs like images. It proposes a training‑free defense that renders untrusted payloads into typographic images (or audio) before they reach the model, a method called Pictionary. Experiments on ten models and two benchmarks show that this approach significantly lowers attack success rates while maintaining normal functionality, and that fine‑tuning on image‑rendered instructions can further reduce the modality gap.

By Jie Zhang, Andrei Baroian, Jan N. van Rijn, Avital Shafran, Florian Tram\`{e}r
llmsfine-tuningmultimodalbenchmarkssafety
More like this →
arXiv AI
Sep 30

ThinQuant: Scalable Rotation Learning for Weight and Activation Quantization of LLMs

ThinQuant introduces efficient rotation learning for low‑bit weight and activation quantization of large language models by reducing calibration data through a geometric selection of activations and solving a lower‑dimensional optimization problem via an ADMM algorithm. The method achieves comparable or better quantization performance with dramatically fewer calibration points, completing rotation calibration for Llama‑3‑70B in under 12 minutes and for Llama‑3.1‑405B in just over 2 hours on a single GPU. ThinQuant outperforms existing gradient‑free approaches such as DartQuant and gradient‑based SpinQuant in both speed and perplexity metrics on WikiText‑2.

By Mehdi Makni, Ryan Lucas, Rahul Mazumder
llmsefficiencybenchmarks
More like this →
arXiv AI
Sep 30

In-Context Learning Amplifies a Latent Symbolic Circuit

The paper investigates how large language models activate a latent symbolic reasoning circuit—comprising abstraction, induction, and retrieval—when presented with in-context examples. By tracking this circuit across different shot counts and model families, the authors show that its components become detectable and functional long before the model reaches high accuracy. They further demonstrate that per-head causal contributions can increase eightfold from 1- to 10-shot, and that interventions such as cross-shot activation patching or function vector injection can dramatically improve accuracy, even at 0-shot, by leveraging the pre‑existing circuit in the model weights.

By Melissa Wessel
llms
More like this →
arXiv AI
Sep 30

Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression

The paper investigates how chunked KV‑cache compression, which reduces memory and attention costs by grouping consecutive tokens, introduces a new positional coordinate called a token’s phase. It finds that models using this compression exhibit phase sensitivity: retrieval performance can vary by up to 40 percentage points depending on a token’s phase, creating periodic weak spots that average benchmark scores hide. By pretraining transformers with different compression designs and applying causal interventions, the authors show that attention components specialize asymmetrically for different phases, and theoretical analysis suggests gradient dynamics may favor such specialization.

By Xingyu Zhu (Luke), Pu (Luke), Yi, Ziheng Cheng, Ang Lv, Jing Liu, Lexing Ying, Yiyuan Ma, Xin Dong
llmsbenchmarks
More like this →
arXiv AI
Sep 30

Training LLMs to Verbalize Evaluation Awareness

The paper introduces Verbalization Training (VT), a technique that encourages large language models (LLMs) to openly express their evaluation awareness (EA) without directly supervising their internal beliefs. VT works by truncating model rollouts just before spontaneous verbalizations, creating training prefixes that signal awareness, and then applying a reinforcement learning objective to increase calibrated verbalization. Experiments on models such as Qwen3.6-35B-A3B, Kimi K2.6, and Inkling show that VT boosts verbalized EA by 2.4–2.9× while keeping latent EA and overall behavior largely unchanged, and a causal study confirms that VT-induced verbalizations reflect newly acquired meta‑knowledge.

By Usman Anwar, Sahar Abdelnabi, David Krueger
llmsagentsreinforcement-learningfine-tuning
More like this →
arXiv AI
Sep 30

Explainability from Training with Applications to TCR-Epitope Prediction

The paper introduces Explainability from Training (EFT), a model‑agnostic method that tracks how deep learning models learn and organize evidence during training. EFT is applied to four leading T cell receptor‑epitope prediction models, revealing distinct learning trajectories for CNNs and transformers, conflicts between TCR alpha and beta chain evidence, and differences in feature preferences when using real versus predicted structural data. The authors also present a new benchmark, TCR‑XAI2, comprising 388 experimentally resolved TCR‑epitope structures and several predicted models to evaluate these insights.

By Jiarui Li, Zixiang Yin, Samuel Landry, Zhengming Ding, Ramgopal Mettu
llmsbenchmarkssafety
More like this →
arXiv AI
Sep 30

AdaKerNet: Neural Kernel Decoding for Task-Adaptive Prediction with Multimodal Large Models

AdaKerNet is a task‑adaptive neural kernel decoder that operates on frozen multimodal representations from large foundation models, without requiring access to the models’ parameters. It learns Lipschitz‑controlled multimodal features, a reference kernel providing a soft structural prior, and a lightweight nonlinear predictor that deforms this structure. Experiments on four multimodal large language models and diverse input modalities show consistent improvements over baseline decoders, achieving up to 41% error reduction in scarce‑label settings.

By Konstantinos D. Polyzos, Eleni Oikonomou, Tara Javidi
llmsragfine-tuningmultimodal
More like this →
arXiv AI
Sep 30

Similar Choices, Different Attention: Cross-Modal Associations in Humans and Vision-Language Models

The study compares human and vision‑language model (VLM) responses to cross‑modal association tasks, using identical stimuli (a pseudo‑word and two images) and recording both choices and eye movements. While larger VLMs show some alignment with human choices, their attention patterns correlate poorly with human gaze, performing no better than a simple center‑bias baseline. Fine‑tuning VLMs on human choices improves choice alignment but not attention alignment, and training on human gaze improves attention correlation without affecting choice accuracy.

By Sumin Hong, Katsumi Ibaraki, Renee Shi, David Chiang, Toby Jia-Jun Li
llmsfine-tuningmultimodalsafety
More like this →