Tootfinder

Events in LA remind us that in times past, when the state uses excessive force against its own citizens, it ultimately was on the wrong side of history. In the UK, we too have a shady past of parallel events and movements: #Levellers #Chartists

One thing I've noticed about the coverage of the LA protests is the differing use of "non-lethal" and "less-lethal" in describing weapons being used by the state against protesters.
And in this case I think the liability-dodging term "less lethal" is actually more useful, and should be incorporated into style-guides.

This https://arxiv.org/abs/2506.05376 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

A Red Teaming Roadmap Towards System-Level Safety
Large Language Model (LLM) safeguards, which implement request refusals, have become a widely adopted mitigation strategy against misuse. At the intersection of adversarial machine learning and AI safety, safeguard red teaming has effectively identified critical vulnerabilities in state-of-the-art refusal-trained LLMs. However, in our view the many conference submissions on LLM red teaming do not, in aggregate, prioritize the right research problems. First, testing against clear product safety …

Democratic governments* should ALL be against AI - if not for the obvious moral issues, for the fact that
LLMs *pay no taxes*.
When you fire somebody making €60k/year in Germany and replace it with an LLM subscription, you're effectively removing 25K or so from state coffers (because, c'mon, you don't believe OpenAI et al are going to pay taxes, are you?).
* not kleptocracies, of course

"The good news is, those same software workers can still influence the state of the world in a positive light. In the pages of this magazine, now celebrating its fifth year of existence, we have tried to highlight the various ways they can do this:
By promoting a healthy dialogue between management and software engineers.
By showing empathy with one another and with mankind.
By unionizing and fighting against atrocious work conditions."

At the risk of stating the obvious, please do state the obvious! In two separate meetings this week it seemed like someone was arguing against a proposed change and in both cases, they were actually very supportive of the idea, and had already moved on to talk about what else could be done. Stop. Express your support (you can thank the suggester if you like!) so that it's clear, and THEN you can build on the original suggestion.
In both cases, we nearly ended up not doing the original thing!

This https://arxiv.org/abs/2409.03788 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

HSF: Defending against Jailbreak Attacks with Hidden State Filtering
With the growing deployment of LLMs in daily applications like chatbots and content generation, efforts to ensure outputs align with human values and avoid harmful content have intensified. However, increasingly sophisticated jailbreak attacks threaten this alignment, aiming to induce unsafe outputs. Current defense efforts either focus on prompt rewriting or detection, which are limited in effectiveness due to the various design of jailbreak prompts, or on output control and detection, which a…

If you want to show solidarity with Los Angeles, the best thing to do is organize protests against ICE tonight/tomorrow in your own cities.
Search the web for ICE rapid response networks in your own city / state and sign up. Help your neighbors.
https://bsky.app/profile/chadloder.bsk

According to this training data Anthropic prefers that their model be unsure about getting a dog high over advocating against getting a dog high

This https://arxiv.org/abs/2503.10349 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…

Autonomous Robotic Radio Source Localization via a Novel Gaussian Mixture Filtering Approach
This study proposes a new Gaussian Mixture Filter (GMF) to improve the estimation performance for the autonomous robotic radio signal source search and localization problem in unknown environments. The proposed filter is first tested with a benchmark numerical problem to validate the performance with other state-of-the-practice approaches such as Particle Filter (PF) and Particle Gaussian Mixture (PGM) filters. Then the proposed approach is tested and compared against PF and PGM filters in real…

This seems bad. We just learned in the past week via @… that ICE agents are searching Flock cameras nationwide, and so did a Texas cop trying to track down a woman who got an out-of-state abortion. Now, without any guardrails against that, a crypto billionaire wants $9m more of these cameras in SF.
There's an ordinance to accept the gift that will…

Turkish authorities ordered the detention of several opposition party members in Istanbul
and raided opposition-run municipalities on Saturday, state media said,
part of a widening legal crackdown against the opposition and city's jailed mayor.
Istanbul Mayor Ekrem Imamoglu, who is also President Tayyip Erdogan's main political rival and leads him in some polls, was jailed in March pending trial over charges of corruption and aiding a terrorist group, all of which h…

Turkish authorities order detention of dozens of opposition officials in widening crackdown
ANKARA (Reuters) -Turkish authorities ordered the detention of several opposition party members in Istanbul and raided opposition-run municipalities on...

This https://arxiv.org/abs/2505.19887 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

Deconstructing Obfuscation: A four-dimensional framework for evaluating Large Language Models assembly code deobfuscation capabilities
Large language models (LLMs) have shown promise in software engineering, yet their effectiveness for binary analysis remains unexplored. We present the first comprehensive evaluation of commercial LLMs for assembly code deobfuscation. Testing seven state-of-the-art models against four obfuscation scenarios (bogus control flow, instruction substitution, control flow flattening, and their combination), we found striking performance variations--from autonomous deobfuscation to complete failure. We…

First-Passage-Time Asymmetry for Biased Run-and-Tumble Processes
Yonathan Sarmiento, Benjamin Walter, Debraj Das, Samvit Mahapatra, \'Edgar Rold\'an, Rosemary J. Harris
https://arxiv.org/abs/2505.24483

First-Passage-Time Asymmetry for Biased Run-and-Tumble Processes
We explore first-passage phenomenology for biased active processes with a renewal-type structure, focusing in particular on paradigmatic run-and-tumble models in both discrete and continuous state spaces. In general, we show there is no symmetry between distributions of first-passage times to symmetric barriers positioned in and against the bias direction; however, we give conditions for such a duality to be restored asymptotically (in the limit of a large barrier distance) and highlight connec…

This https://arxiv.org/abs/2408.07529 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_qu…

Optimal number of stabilizer measurement rounds in an idling surface code patch
Logical qubits can be protected against environmental noise by encoding them into a highly entangled state of many physical qubits and actively intervening in the dynamics with stabilizer measurements. In this work, we numerically optimize the rate of these interventions: the number of stabilizer measurement rounds for a logical qubit encoded in a surface code patch and idling for a given time. We model the environmental noise on the circuit level, including gate errors, readout errors, amplitu…

Topology meets symmetry breaking: Hidden order, intrinsically gapless topological states and finite-temperature topological transitions
Reja H. Wilke, Henning Schl\"omer, Simon M. Linsel, Annabelle Bohrdt, Fabian Grusdt
https://arxiv.org/abs/2506.03146

Topology meets symmetry breaking: Hidden order, intrinsically gapless topological states and finite-temperature topological transitions
Since the discovery of phase transitions driven by topological defects, the classification of phases of matter has been significantly extended beyond Ginzburg and Landau's paradigm of spontaneous symmetry breaking (SSB). In particular, intrinsic and symmetry-protected topological (SPT) orders have been discovered in (mostly gapped) quantum many-body ground states. However, these are commonly viewed as zero-temperature phenomena, and their robustness in a gapless ground state or against thermal …

This https://arxiv.org/abs/2502.14814 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…

VB-Com: Learning Vision-Blind Composite Humanoid Locomotion Against Deficient Perception
The performance of legged locomotion is closely tied to the accuracy and comprehensiveness of state observations. Blind policies, which rely solely on proprioception, are considered highly robust due to the reliability of proprioceptive observations. However, these policies significantly limit locomotion speed and often require collisions with the terrain to adapt. In contrast, Vision policies allows the robot to plan motions in advance and respond proactively to unstructured terrains with an o…

#Publishers and #Editors looking for guidance on how to #DefendResearch against #censorship

"I don’t believe anyone was thinking about [the] 51st state idea. Truthfully, the intent was to not get political, but to focus on social, cultural, and economic ties."
I hate this nonsense. With great respect and in kindness: Mr. City Councillor in Bellngham needs to put on his big boy pants and admit that yes it is political and yes it is about the 51st state rhetoric and all the scary border stuff from his federal government.
His choice not to admit that and speak against that, and worse, to act like it doesn’t exist, is a further insult and will not help to get BC/Canadians to resume visits.
#CanPoli #CdnPoli #51stState #NoFascism #TheAmericanFascist #Economy #Tourism
https://www.albernivalleynews.com/home2/video-bellingham-letter-of-support-sent-to-bc-communities-8034581