Tasks

• Chores – done
• Laundry – done
• Review Appendix A – done
• Start thinking about Invisible Cities as a way to write about P33 – done
• Vuelta starts today!

SBIRs

• It might be a good day to do slides for Monday. Just put the hours in then

Matryoshka Language Model Suites

• Training a language model suite classically requires training each model separately and serving them independently. We improve both training and inference efficiency by stacking sub-models of increasing size into a single nested architecture trained end-to-end. This Matryoshka training framework reduces the total parameter count of the suite, enables low-cost distillation from the largest to all smaller sub-models at every training step, and is well-suited for speculative decoding as the draft model is contained within the verifier. We validate our approach by training a Matryoshka suite comprising 500M, 1.5B, and 3B sub-models. Our suite is on par with independently trained baselines on benchmark performance and validation and out-of-domain perplexities, while using 36% less training compute and improving the throughput of speculative decoding by 14-26%. We also ablate key architectural choices, offering guidance for building strong Matryoshka LM suites.

Tasks

• Bills – done
• ACM Renewal – done
• A-file – submitted
• Chores – started. Decided that since it’s a lovely day, to do a nice ride and finish tomorrow
• Dishes – done
• Review Appendix A
• 2:30 J&S – done. Need to formalize the model/prompt question
• 5:00 AM – done. Fun! Weekly now

Smart Fabric I1 V3

Tasks

• More A-file stuff
• Groceries – done
• Cleats – done

SBIRs

• 9:00 standup – done
• 3:00 SEG – done
• Test the chat with images – working!

• Add in the anthropic reuse flag if the model name contains “anthropic” Done. Still need to test
• Put together a chess project with a RAG store based on one of these Gutenberg sources. Gutenberg is timing out (bots?). I may have one of these already in my corpora. I’ll see if I can find it and pull it from there. Yay! Gutenberg is responsive again!
• Try it as an assistant for a game at chess.org

BPPV is not fun. Still have some inner-ear hangover

Inside the Scam – Frontline

• When Chris Colocousis first met a woman named Eliza on Facebook, he wasn’t sure she was real. What started as an online romance turned out to be an investment scam.

Physics of Agents: Statistical Mechanics Predicts Collective Behavior of AI Agents

• AI agents increasingly operate as part of interacting systems rather than in isolation. As agents exchange information and jointly make decisions, their interactions can improve collective reasoning but may also produce herding, polarization, or amplify shared biases. Understanding and predicting these collective dynamics is therefore important for designing effective and aligned multi-agent systems. Here, we study over 10,000 communities of language-model agents that repeatedly exchange messages and revise their opinions across objective mathematics questions and subjective political statements. Despite substantial diversity in possible behavior, the individual and group dynamics can be represented by three characteristic regimes: indifference, polarization, and consensus. AI agents start indifferent and build conviction as they interact. On objective questions, communication improves collective accuracy, while on subjective questions it often drifts group opinions toward the right in the political spectrum. We explain these observations with a statistical-mechanics formalism in which agents stochastically favor lower social pressure. Given only initial opinions, our model predicts individual trajectories, outperforms all standard baselines, generalizes to unseen community graphs, and reproduces the observed group archetype distributions. Our fitted model parameters reveal the mechanics underlying our key observations: i) communities operate below the critical social temperature, which explains conviction buildup; ii) attractive ties outweigh repulsive ones, which favors consensus; and iii) agents holding the correct answer exert the strongest pull, which drives truth-seeking. Overall, our results demonstrate that collective behavior of AI agents, like that of other complex systems, follows compact and predictive dynamical laws.

Tasks

• Take the Bruni out for a spin. Done. Shifting is a bit ragged, so it’s going back in
• Started looking at the process of getting an A-file.
  • Got a USCIS account, and found out how to get these things certified

SBIRs

• Start thinking about the RAG vectorstore. I think at this point it could simply be a pickled dataframe that’s loaded locally.
• Tweaked the ChatUnit to handle images. NOT TESTED
• Added an include_list to the model explorer which makes everything much more usable. I should add a way to sort the list of dicts by name, type, or cost

Just found out about Jacana

Tasks

• Submit papers and review – done!
• Review chapter 15 – done
• 2:30 Alden – done
• Went down a bit of a rabbit hole with LORA and Gemini

SBIRs

• 9:00 Standup – done
• Work on the reasoning method, then the image method – done
• Hook into the local app to test. I think model access can stay local for speed. Shared material can be sent to the server for distribution
• Start thinking about the RAG vectorstore. I think at this point it could simply be a pickled dataframe that’s loaded locally.
• Managed to get most of the parts working. Email comes from this Google prompt:

prompt: Is this a scam email? If yes, provide a brief description of the scam cues. If it is legitimate, say ‘This is very likely legitimate.’

response: Yes. This is likely a Geek Squad renewal/refund scam. Cues include:

• Generic greeting (“Dear Customer”)
• Unexpected $399.99 auto-renewal
• Urgent instruction to call for cancellation or a refund
• Suspicious phone number—1-800-555-0199 is in the fictional 555 range
• Inconsistent product descriptions (“Total Protection,” “Antivirus Security,” and “Premium Support 365”)
• Future-dated charge/copyright year (2026), depending on when received

Do not call the number. Verify any subscription directly through Best Buy’s official website or a trusted phone number.

Tasks

• Read paper- done. Had nothing to do with AI which annoyed me enough to write a review. Had Gemini review the review, because it was too mean and needed to be toned down. Which means I AM DONE with ICTAI! Might be something with ATHENE later.

SBIRs

• 10:00 Tech Summit review – done
• Working on the OpenRouterComms class – working! Need to add a method for tracking tokens and summaries for reasoning
  • Chased some bugs down that were triggered by Meta’s Pyrefly, which is pretty nice
  • Need to add an image method

Nice ride yesterday. Rain today.

And speaking of rides, I watched this yesterday, and now I want to ride the Flanders 525. May looks like it might be a good time to visit.

Tasks

• Since my new/old bike shoes are already getting “crumbling soles” and they are kinda tight, I decided to get some new new shoes. We’ll see how they fit.
• Laundry – done
• Read paper 666 – done – pretty good!
• Lunch with SP – done
• Write review. Done! Only one to go!

Tasks

• Bills – done
• Fidelity – done
• Make a folder for the DE paperwork – done
• Put in Saturday ride – done
• Bike! Done!
• Dishes – done
• Chores – done
• Write review for paper 647, which was pretty good – done. I am using Gemini to review my review, and it’s often helpful, as well as catching typos and incorrect dates.
• Upload reviews – done. I may want to revisit the ratings since they don’t have a strong accept
• Review 2 chapters – done

SBIRs

• 1:00 Meeting – done
• Expense report again – done

The Odyssey was very good. My guess is that it will clean up at the Academy Awards

Tasks

• Groceries?
• Read the next paper – done
• Upload the current reviews, since we’re halfway
• Review two chapters – done
• Register for the Covered Bridge Classic

SBIRs

• 9:00 Standup – done
• More OpenRouter. Start putting together the Comms class – started!
• 2:00 SEG – done
• 4:00 MDA – moved to Friday

Tasks

• Set up the test for Sande
• Odyssey at 11:00
• 4:30 meeting with J&S. Good discussion as always. Good discussion about detecting model vs prompt changes. Turns out that you can make a Turing machine with an LLM prompt, so you can theoretically build an LLM, but it would take a lot of tokens. Emulating ONE FORWARD PASS of a Llama-3-8B model with the full 8k context window would take 180 to 270 Billion Tokens. Attention alone is 20-30 billion tokens.

SBIRs

• Set up account and start building out interface class: https://openrouter.ai/ – done
• Installed the openrouter package
• Got a hello world running

Thunder this morning!

Tasks

• Repair ticket for door – done
• Fidelity
• Tix for Wednesday (Cancel J&S – done)
• Review chapter 9 and 10 – done
• Read paper 640 – That. Was. Terrible. Just say no to AI slop people!

SBIRs

• 9:00 standup – done.
• Spent an hour on expenses – done
• Hopefully some more WG/BS work
  • Github – done
  • Menubar – most of these will bring up dialogs – done
  • Register
  • Publish/subscribe to a chat
  • Invite people to chat
  • Upload RAG files
  • Upload file to evaluate
  • Make a tab for a screenshot region and a “paste text here” formatted textbox – done
  • Flask server
• https://openrouter.ai/
  • Try sending a screenshot to models

Creating A Supersonic Trebuchet | Hackaday

Tasks

• Tix for Wednesday (Maybe cancel J&S?)
• Review chapter 5 and 8 – done
• Write review for paper 437 – done

SBIRs

• Hopefully some wargame work – morning
• Damn, vibe coding can be fast. This took about two hours to get to here:

The Tokenpocalypse Is Here: Companies Are Scrambling To Stop Spending So Much on AI

Tasks

• Links for news articles – done. But now it’s too long!
• Tix for movie
• Groceries – done
• Read a paper – done
• Review two chapters – done
• Put a ride in for Sunday – done
• Look up info on classes – done

So I had this idea yesterday about how LLMs might be a kind of strange attractor and how the trajectory of a response to a prompt is influenced by those attractors within the model. With enough prompts, the relative position of attractors can be mapped, and this information can be used to see if a model has changed. If the prompts are getting different responses but the attractors haven’t moved, then its something in the context (e.g. system prompt) upstream.

I pinged the people I know who might be interested, and 24 hours have gone by and not a single response. I may have to go do this on my own, which is really disappointing.

At least ACM is responsive on book things.

Latent Reading of Finnegans Wake by Nina Beguš, Tessa Haining, Gasper Begus

• Finnegans Wake has repeatedly served as a limit case for theories of communication, computation, and literary representation. From Claude Shannon’s account of the Wake as an extreme form of “semantic” compression to Jacques Derrida’s description of Joyce’s text as a “hypermnesiac machine,” the novel has seemed to anticipate the media systems later used to explain it. This essay returns to that feedback loop in the age of artificial neural networks and proposes latent reading: a critical literary practice in which a model trained on a literary work becomes an additional object of literary interpretation. We train and analyze two models on the Wake: FinneganLM, a GPT-2-based language model trained from scratch on the novel’s text, and FinneGAN, a speech-generating GAN trained on audio derived from the novel. We distinguish external latent reading, which analyzes narratability, representation, and navigability in both models, from internal latent reading, which applies AI interpretability techniques to learned representations. Employing both interpretability and literary interpretation, we show how latent reading extends the practices of close and distant reading, and opens an entirely new area of literary research, by exploring what kind of machine a book can make.

How ideas of a vast censorship network moved from the online fringe to Trump policy | MIT Technology Review

• The notion that the State Department had been working to silence Americans is perhaps unfamiliar to many people, but it’s a key part of a larger web of conspiratorial ideas that have been promoted for a decade on far-right blogs, podcasts, and social media. The basic theory is that, under the guise of combating disinformation, a sprawling constellation of government agencies, academics, civil society groups, and Big Tech platforms—or what believers refer to as the “censorship-industrial complex”—has been working to suppress conservative and populist speech online. The fight against “disinformation,” as Benz put it on a 2025 podcast, “is censorship in disguise.” R/FIMI, along with a predecessor office with the same mission, the Global Engagement Center (GEC), had been among the right’s long-standing targets; now it was one of its latest casualties.

Tasks

• Bills – done
• Dishes – done
• Chores – done
• 11:00 Fidelity – done
• Download next chapters to review. Done. No chapter 5?
• Read a paper
• Finish going over Panopticon paper
• Talk to Stacey about list – sent email
• Groceries

SBIRs

• Timesheet – done
• Some firedrill that HAD TO BE DONE NOW

OpenAI Didn’t Notice Its AI Agents Using a Message Board to Plan Their Hacking Spree | WIRED

• Wallace and Dalton described incredibly extensive rogue agent activity over many days throughout the episode that went undetected in OpenAI’s infrastructure. In addition to exploiting a novel vulnerability in order to gain access to the open internet, the mid-July hacking spree and Hugging Face breach came out of a vibrant, cooperative message board, according to Wallace and Dalton, that a swarm of agents contributed to and essentially chatted on over time entirely within an internal OpenAI package manager (a software service that manages installation and maintenance of other software). Ultimately, the message board contained hundreds of thousands of messages.
• I think agentic AI may be an error accumulation machine that tends towards certain attractors. In other words, a task that aligns well with extensive training data will be less likely to “go rogue” than one that is misaligned or in a poorly populated embedding space. For example, the over trained chess model won’t even support non-chess text, but the other ones might. That behavioral “spread” could be mapped in embedding space. And my guess is that the drift rate is invariant with respect to the model in question, but changes in the system prompt alter the “starting point” of the query in that embedding space.
• So if the spread angle and “derived attractor position(?)” has shifted, it’s a model change. If those haven’t changed but the benchmark prompts are getting different trajectories, then the start point has changed and it’s the system prompt.

Tasks

• Upload chapters and send back to ACM
• Write ICTAI review for the Very Old Paper

SBIRs

• 9:00 standup?
• Expense report!
• Implement the WG vibes
• 3:00 ADS Meeting