Tootfinder

When Truthful Representations Flip Under Deceptive Instructions?
Xianxuan Long, Yao Fu, Runchao Li, Mu Sheng, Haotian Yu, Xiaotian Han, Pan Li
https://arxiv.org/abs/2507.22149 h…

When Truthful Representations Flip Under Deceptive Instructions?
Large language models (LLMs) tend to follow maliciously crafted instructions to generate deceptive responses, posing safety challenges. How deceptive instructions alter the internal representations of LLM compared to truthful ones remains poorly understood beyond output analysis. To bridge this gap, we investigate when and how these representations ``flip'', such as from truthful to deceptive, under deceptive versus truthful/neutral instructions. Analyzing the internal representations of Llama-…

Quick bit of emergency repairing for my boy’s fav combats!
Added a patch behind what is a very threadbare area

AI startup Aurasell, which aims to automate sales tools built on top of CRM software like Salesforce, emerges from stealth with $30M seed led by Next47 (Geoff Weiss/Business Insider)
https://www.businessinsider.com/ai-sales-startup-aurasell-raise…

AI startup Aurasell raised $30M seed in 28 hours to take on Salesforce
AI sales startup Aurasell is emerging from stealth with a $30 million seed round that its founders said closed in roughly a day.

Has there ever been a scene of witches flying around on brooms in a movie that has realistic acceleration physics etc?
I'm picturing a pair of witches in tight formation at high altitude making a steep dive to pick up speed, then pull out at treetop height, a high G turn to line up for a low pass over a bunch of trick-or-treaters, then maybe getting a ping on their RWR when some grownups come into view.

EU advisors propose turning CO2 into a tradeable commodity in new blueprint. #climatechange #climatesolutions #climate

Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
Yao Fu, Xianxuan Long, Runchao Li, Haotian Yu, Mu Sheng, Xiaotian Han, Yu Yin, Pan Li
https://arxiv.org/abs/2508.19432

Quantized but Deceptive? A Multi-Dimensional Truthfulness Evaluation of Quantized LLMs
Quantization enables efficient deployment of large language models (LLMs) in resource-constrained environments by significantly reducing memory and computation costs. While quantized LLMs often maintain performance on perplexity and zero-shot tasks, their impact on truthfulness-whether generating truthful or deceptive responses-remains largely unexplored. In this work, we introduce TruthfulnessEval, a comprehensive evaluation framework for assessing the truthfulness of quantized LLMs across thr…

Incentive-Aligned Multi-Source LLM Summaries
Yanchen Jiang, Zhe Feng, Aranyak Mehta
https://arxiv.org/abs/2509.25184 https://arxiv.org/pdf/2509.25184

Incentive-Aligned Multi-Source LLM Summaries
Large language models (LLMs) are increasingly used in modern search and answer systems to synthesize multiple, sometimes conflicting, texts into a single response, yet current pipelines offer weak incentives for sources to be accurate and are vulnerable to adversarial content. We introduce Truthful Text Summarization (TTS), an incentive-aligned framework that improves factual robustness without ground-truth labels. TTS (i) decomposes a draft synthesis into atomic claims, (ii) elicits each sourc…

Feeding Two Birds or Favoring One? Adequacy-Fluency Tradeoffs in Evaluation and Meta-Evaluation of Machine Translation
Behzad Shayegh, Jan-Thorsten Peter, David Vilar, Tobias Domhan, Juraj Juraska, Markus Freitag, Lili Mou
https://arxiv.org/abs/2509.20287

Feeding Two Birds or Favoring One? Adequacy-Fluency Tradeoffs in Evaluation and Meta-Evaluation of Machine Translation
We investigate the tradeoff between adequacy and fluency in machine translation. We show the severity of this tradeoff at the evaluation level and analyze where popular metrics fall within it. Essentially, current metrics generally lean toward adequacy, meaning that their scores correlate more strongly with the adequacy of translations than with fluency. More importantly, we find that this tradeoff also persists at the meta-evaluation level, and that the standard WMT meta-evaluation favors adeq…

Turning compute into a tradable commodity could fuel the next stage of the AI boom, just like oil futures and spectrum auctions unlocked waves of investment (Felix Salmon/Bloomberg)
https://www.bloomberg.com/news/articles/20