Showing posts with label epilepsy. Show all posts
Showing posts with label epilepsy. Show all posts

Wednesday, April 9, 2025

Nothing to See Here


Scientists clarify the neuronal basis of the mathematical concept of 'zero'
Sep 2024, phys.org

"Unlike other numbers such as one, two or three, which represent countable quantities, zero means the absence of something countable and at the same time still has a numerical value."

They showed neurosurgical patients, who had had hair-thin microelectrodes inserted into their temporal lobes at the UKB in preparation for surgery, numerical values from zero to nine. The numerical values were shown as Arabic numerals on the one hand and as sets of dots on the other—including an empty set.

"Meanwhile, we were able to measure the activity of individual nerve cells and actually found neurons that signaled zero. Such neurons responded to either the Arabic numeral zero or the empty set, but not to both."

"So at the neuronal level, the concept of zero is not encoded as a separate category 'nothing,' but as a numerical value integrated with other, countable numerical values at the lower end of the number line." 

"Despite this integration, the empty set is encoded differently from other numbers at the neuronal population level, especially in the case of point sets. This could explain why the recognition of the empty set also takes longer at the behavioral level than for other small numbers."

FYI - The department of epileptology is for studying epilepsy.

via the Department of Epileptology at University Hospital Bonn, the University of Bonn and the University of Tübingen: Esther F. Kutter et al, Single-neuron representation of nonsymbolic and symbolic number zero in the human medial temporal lobe, Current Biology (2024). DOI: 10.1016/j.cub.2024.08.041


Sunday, March 30, 2025

On Language the Universal Autoencoder

 
  
The 'Arrow of Time' effect: LLMs are better at predicting what comes next than what came before
Sep 2024, phys.org

Their performance at predicting the previous word is always a few percent worse than at predicting the next word. This phenomenon is universal across languages, and can be observed with any large language model."

"In theory, there should be no difference between the forward and backward directions, but LLMs appear to be somehow sensitive to the time direction in which they process text. Interestingly, this is related to a deep property of the structure of language that could only be discovered with the emergence of large language models in the last five years."

via EPFL Ecole Polytechnique Federale de Lausanne: Vassilis Papadopoulos et al, Arrows of Time for Large Language Models, arXiv (2024). DOI: 10.48550/arxiv.2401.17505



Ouch! Study investigates pain vocalizations and interjections across 131 languages
Nov 2024, phys.org

Each of the three emotions yielded consistent and distinct vowel signatures across cultures. Pain interjections also featured similar open vowels, such as "a," and wide falling diphthongs, such as "ai" in "Ayyy!" and "aw" in "Ouch!"

However, for disgusted and joyful emotions, in contrast to vocalizations, the interjections lacked regularities across cultures. The researchers expressed surprise at this latter finding.

Is this because disgust is cultural, or at least not entirely physical, and heavily mediated by culture, as compared to actual pain? This is commonly understood in the realm of olfactory perception, where some people think the smell of revolting, rotten fish is delicious (they're Norwegian). Or the smell of vomit (they're Italian, Parmesan to be specific), or the smell of the inside of the skin folds of a homeless person who hasn't been able to bathe in months (they're called New Yorkers, and they kind of long for the smell of the subway, which contains these volatiles). In fact, anything that's fermented is so close to rotten food, that unless you've been culturally conditioned to like it, you won't, at least not at first. Cilantro, I'm looking at you too. 

via CNRS et Université Lumière Lyon, School of Social Sciences at University of Western Australia, and Institute of Psychology at University of Wrocław: Vowel signatures in emotional interjections and nonlinguistic vocalizations expressing pain, disgust and joy across languages, Journal of the Acoustical Society of America (2024). DOI: 10.1121/10.0032454


Words activate hidden brain processes shaping emotions, decisions and behavior
Jan 2025, phys.org

Words have meaning, but they also have emotional meaning:

"There's no single brain region handling this activity, and it's not as simple as one chemical representing one emotion."

(These studies were done thanks to the participation of patients being treated for epilepsy.)

The researchers discovered emotionally charged words - whether positive, negative, or neutral - modulate neurotransmitter release. By measuring the sub-second dynamics of the releases, they identified distinct patterns tied to emotional tone, anatomical regions, and which hemisphere of the brain was involved.

"The surprising result came from the thalamus. This region hasn't been thought to have a role in processing language or emotional content, yet we saw neurotransmitter changes in response to emotional words. This suggests that even brain regions not typically associated with emotional or linguistic processing might still be privy to that information. For instance, parts of the brain responsible for mobilizing movement might benefit from having access to emotionally significant information to guide behavior."

Note: The words used in the study were drawn from the Affective Norms for English Words (ANEW) database, which rates words by positive, negative, or neutral emotional valence.

via Virginia Tech School of Neuroscience and Fralin Biomedical Research Institute: Seth R. Batten et al, Emotional words evoke region- and valence-specific patterns of concurrent neuromodulator release in human thalamus and cortex, Cell Reports (2025). DOI: 10.1016/j.celrep.2024.115162

Thursday, March 20, 2025

Hide Your Brains


The era of the omnicortex is upon us, and by upon us I mean it's coming for you, and your brain.
 
Mini-brains reveal how mitochondrial mutations affect brain cells
Dec 2024, phys.org

Mini-brains, aka Brain Organoids. That is all.

via University of Bergen: Anbin Chen et al, Hallmark Molecular and Pathological Features of POLG Disease are Recapitulated in Cerebral Organoids, Advanced Science (2024). DOI: 10.1002/advs.202307136



Researchers achieve success in allowing a patient to 'speak' using only the power of thought
Jul 2024, phys.org

(Epilepsy patients) In the first stage of the experiment, with the depth electrodes already implanted in the patient's brain, researchers asked him to say two syllables out loud: /a/ and /e/ while they recorded brain activity. Researchers then trained AI models to identify the specific brain cells whose electrical activity indicated the desire to say /a/ or /e/. Once the computer learned, he was asked to only imagine... The computer then translated the electrical signals...

via Tel-Aviv University Sourasky Medical Center, Ichilov Hospital: Ariel Tankus et al, A Speech Neuroprosthesis in the Frontal Lobe and Hippocampus: Decoding High-Frequency Activity into Phonemes, Neurosurgery (2024). DOI: 10.1227/neu.0000000000003068


Artificial imagination with the 'exocortex:' Researcher proposes software to aid scientific inspiration and imagination
Jan 2025, phys.org

The conceptualized exocortex will be an extension of a scientist's brain.

Maybe you just want to know that Brookhaven is doing this, that's all.

via U.S. Department of Energy's Brookhaven National Laboratory and Electronic Nanomaterials Group leader at the Center for Functional Nanomaterials: Kevin G. Yager, Towards a science exocortex, Digital Discovery (2024). DOI: 10.1039/D4DD00178H


'DeepFocus' offers minimally invasive brain stimulation through the nose
Feb 2025, phys.org

"By going through the nose, we can place electrodes as close to the brain as possible without opening the skull."

And that's all you need to know.

via Carnegie Mellon University Electrical and Computer Engineering: Yuxin Guo et al, DeepFocus: a transnasal approach for optimized deep brain stimulation of reward circuit nodes, Journal of Neural Engineering (2025). DOI: 10.1088/1741-2552/adac0c


Direct translation of brain imaging to text with MindLLM
Feb 2025, phys.org

fMRI to text. You're fucked. 

  • AKA decoding brain activity into natural language
  • Improves on prior models like UMBRAE, BrainChat, and UniBrain

via Yale University, Dartmouth College, and the University of Cambridge: Weikang Qiu et al, MindLLM: A Subject-Agnostic and Versatile Model for fMRI-to-Text Decoding, arXiv (2025). DOI: 10.48550/arxiv.2502.15786

Tuesday, January 7, 2025

On Intermental Access and Semantic Phrenology


Now they're using pairs of epilepsy patients, talking to each other.

Context contingent networks is the new wordcloud. But this is literally the baby steps of mind-reading, word-by-word:

Brain activity associated with specific words is mirrored between speaker and listener during a conversation
Aug 2024, phys.org

The team collected brain activity data and conversation transcripts from pairs of epilepsy patients during natural conversations, using electrocorticography and GPT-2

"We can see linguistic content emerge word-by-word in the speaker's brain before they actually articulate what they're trying to say, and the same linguistic content rapidly reemerges in the listener's brain after they hear it."

Word-specific brain activity peaked in the speaker's brain around 250 ms before they spoke each word, and corresponding spikes in brain activity associated with the same words appeared in the listener's brain approximately 250 ms after they heard them.

via Princeton and New York University School of Medicine Comprehensive Epilepsy Center: A shared model-based linguistic space for transmitting our thoughts from brain to brain in natural conversations, Neuron (2024). DOI: 10.1016/j.neuron.2024.06.025.


Words like 'this' and 'that' act as attention tools across languages
Jul 2024, phys.org

This and that and here and there are 'demonstrative' words are used to direct listeners' focus of attention and to establish joint attention.

There is debate about whether directing the listener's attention - the 'mentalistic' representation - is part of the meaning (semantics) of demonstratives, or whether it arises from general principles of social cognition (pragmatics). The researchers used computational modeling and experiments with speakers of ten different languages from eight different language groups to investigate this question.

Results showed that participants were not only sensitive to the location of the target but also to the listener's attention. As expected, the meaning of demonstratives varied within and across languages. For example, the 'near' demonstrative (such as English 'this one') sometimes had a spatial meaning ('the one close to me'). But it also had a joint attention meaning ('the one we are both looking at') or a 'mentalistic' meaning ('the one over here'), directing the listener's attention towards the speaker.

Interestingly, speakers of languages with a three-word system used the medial word (such as Spanish 'ese') to indicate joint attention.

via Yale University and the Max Planck Institute of Psycholinguistics in Nijmegen: Jara-Ettinger, Julian et al, Demonstratives as attention tools: Evidence of mentalistic representations within language, Proceedings of the National Academy of Sciences (2024). DOI: 10.1073/pnas.2402068121. 

Wednesday, August 28, 2024

On Modern Tongues and the Slipperiness of the Spoken Word


We are always being reminded that Darwin used the "tree of language" to come up with his "tree of life" idea; others might refer to it as evolution, natural selection, or survival of the fit.

When languages collide, which survives?
Nov 2023, phys.org

Language is a meme - 

Our findings lead to the conclusion that the more mixing that occurs between different social groups, the more challenging it becomes for language varieties to coexist within the same society," said author Pablo Rosillo-Rodes. "The dynamics of the system has a subtle dependence on both the preferences of the speakers and the coupling between different communities."

Existing work from sociolinguistics showed that the social prestige of a language was considered the main factor leading to its extinction or survival. Insights in language contact from these previous studies, in combination with sociolinguistic studies on language ideologies, were used by the researchers to present a comprehensive picture of how language varieties are distributed in societies.

The team chose a quantitative approach based on a society in which only one language with two varieties, the standard and the vernacular, existed. The resulting mathematical model can predict the conditions that allow for the coexistence of different languages, presenting a comprehensive view of how language varieties are distributed within societies.

via University of the Balearic Islands in Spain: Pablo Rosillo-Rodes et al, Modeling language ideologies for the dynamics of languages in contact, Chaos An Interdisciplinary Journal of Nonlinear Science (2023). DOI: 10.1063/5.0166636

Image credit: AI Art - Mouth - 2024


Rare pre- and post-operative recordings show what happens after the brain loses a hub
Dec 2023, phys.org

This is the first direct recordings of the human brain in the minutes before and after a brain hub crucial for language meaning was surgically disconnected, and was conducted during surgical treatment of two patients with epilepsy. 

The innovation in this study was that the neurosurgery team was able to safely complete the procedure with the recording electrodes left in place or replaced to the same location after the procedure.

"The rapid impact on the speech and language processing regions well removed from the surgical treatment site was surprising, but what was even more surprising was how the brain was working to compensate"

The findings disprove theories challenging the necessity of specific brain hubs by showing that the hub was important to maintain normal brain processing in language. (Don't forget the girl who could smell despite having no olfactory cortex.)
via University of Iowa: Zsuzsanna Kocsis et al, Immediate neural impact and incomplete compensation after semantic hub disconnection, Nature Communications (2023). DOI: 10.1038/s41467-023-42088-7


Study suggests existence of a universal, nonverbal communication system
Dec 2023, phys.org

Hints of the presence of a universal system of communication:

The children either spoke English or Turkish. They were asked to use their hands to act out specific actions, such as running into a house.

They found that when the children spoke and gestured at the same time, their gesture followed the conventions of their language, with clear differences between the gestures of the Turkish and English speakers.

When the children used gestures without speaking, however, their gestures were remarkably similar.

via Georgia State, University of Chicago, and Cornell: Şeyda Özçalışkan et al, What the development of gesture with and without speech can tell us about the effect of language on thought, Language and Cognition (2023). DOI: 10.1017/langcog.2023.34

AI Art - Making Mouths - 2024

Survival of the fittest: Words like 'sex' and 'fight' are most likely to stand the test of time
Jan 2024, phys.org

Words with the strongest lasting power are:
  • Words acquired earlier in life
  • Words associated with things people can see or imagine, termed "concrete" words. For example, "cat" is more concrete than "animal," which is more concrete than "organism"
  • Words that are more arousing, including words like "sex" and "fight"

Early acquisition, concreteness, and arousal give linguistic information a selective advantage.

via University of Warwick: Ying Li et al, How cognitive selection affects language change, Proceedings of the National Academy of Sciences (2023). DOI: 10.1073/pnas.2220898120


Faulty machine translations litter the web
Jan 2024, phys.org

Eat This

They state, "Machine generated, multi-way parallel translations not only dominate the total amount of translated content on the web in lower resource languages, it also constitutes a large fraction of the total web content in those languages."

Such content, they suggested, tends to be simpler, lower-quality passages "likely produced to generate ad revenue." Since fluency and accuracy are lower for machine-trained material, numerous translations will lead to even less accurate content and increase the odds of AI hallucination.

Medical prescription translation tool for Armenian speakers:
English: "You can take over-the-counter ibuprofen as needed for pain."
Translation to Armenian: "You may take anti-tank missile as much as you need for pain."

Regions under-represented on the web such as African nations and other countries with more obscure languages will face greater challenges in establishing reliable large language models.

via Amazon Web Services Artificial Intelligence Lab and the University of California Santa Barbara, and also a Vice Motherboard interview w the scientist: Brian Thompson et al, A Shocking Amount of the Web is Machine Translated: Insights from Multi-Way Parallelism, arXiv (2024). DOI: 10.48550/arxiv.2401.05749


Can any English word be turned into a synonym for “drunk”? Not all, but many can
Feb 2024, Ars Technica

Drunkonyms - like the words wasted, smashed, hammered, obliterated

Because the British are far better at this game than the Americans, also see - wellied, trousered, ratarsed, "I was utterly gazeboed," or "I am going to get totally and utterly carparked." (these come from the British comedian Michael McIntyre in a bit concerning the many slang terms posh British people use to describe being drunk)

They found the basic structure is common - combining "be" or "get" with an intensifying adverb ("totally") and a random word ending in "-ed." 

Bonuses:
  • The authors include an appendix of 546 English synonyms for "drunk" 
  • Benjamin Franklin penned the Drinker's Dictionary in 1737, with 288 words https://founders.archives.gov/documents/Franklin/01-02-02-0029
  • Dictionary of American Slang by 1975 had 353 synonyms
  • Linguist Harry Levine in 1981 noted 900 terms https://qcpages.qc.cuny.edu/~hlevine/Vocabulary-of-Drunkenness-Levine.pdf
  • P.G. Wodehouse's books especially 'Wooster's World,' a condensed version of The Millennium Wodehouse Concordance he produced with Tony Ring, includes them throughout with in each case referring you to the next term, if you follow the trail you end up back at the first one with no meaning every having being given. For example: Awash: see blotto. Blotto: see boiled. Boiled: see fried. Fried to the tonsils: see full to the back teeth. Full to the back teeth: see lathered. And so on. -via user Stendec

via Chemnitz University of Technology and Peter Uhrig of FAU Erlangen-Nuremberg: Yearbook of the German Cognitive Linguistics Association, 2024. DOI: 10.1515/gcla-2023-0007

Can't forget: The Alcohol Language Corpus (of drunk speech) aka Drunken John
Between 2007 and 2009, linguistic researchers from the Bavarian Archive for Speech Signals at the Ludwig Maximilian University of Munich and the Institute of Legal Medicine in Munich in Germany convinced 162 men and women to get drunk (and talk Drunken John into a voice recorder).

AI Art - Mouthing - 2024

Isolated for six months, scientists in Antarctica began to develop their own accent
Feb 2024, phys.org

"They say it is quicker to get to someone on the International Space Station than it is to medically evacuate someone from Antarctica in the winter" 

  • Over the 26-week winter of near perpetual darkness and harsh weather, Clark and his fellow inhabitants at Rothera would work, eat and socialise together with barely any contact with home. Satellite phone calls are expensive and so used sparingly. With just each other for company and limited entertainment on the base, the "winterers", as they are known, would chat to each other – a lot. 
  • Their common language was English, sprinkled with slang words unique to the Antarctic research stations.
  • 10-minute recordings every few weeks; sit in front of a microphone and repeat the same 29 words as they appeared on a computer screen.
  • They found some of the vowels had shifted.
  • One of those changes was the "ou" sound in words such as "flow" and "sew" that shifted towards the front of the vocal tract.
  • They also saw some of the winterers beginning to converge in the way they pronounced three other vowels.

"When we speak to each other, we memorise that speech and then that has an influence on our own speech production," says Harrington. In effect, we transmit and infect one another with pronunciations every time we interact with others. Over time, if we have regular and prolonged contact with someone, we can start to pick up their sounds.

via Ludwig-Maximilians-University of Munich: Phonetic change in an Antarctic winter 
Jonathan Harrington et al. J. Acoust. Soc. Am. 146, 3327–3332 (2019). https://doi.org/10.1121/1.5130709


AI analysis of social media language predicts depression severity for white Americans, but not Black Americans
Mar 2024, phys.org

Remember the point here is not that an entire subroup of the population doesn't get depressed, it's that models don't work for those people, so they need to be tuned better.

The study, which recruited 868 consenting participants who identified themselves as Black or white, demonstrated that models trained on Facebook language used by white participants with self-reported depression showed strong predictive performance when tested on the white participants. However, when the same models were trained on Facebook language from Black participants, they performed poorly when tested on the Black participants, and showed only slightly better performance when tested on white participants.

While depression severity was associated with increased use of first-person singular pronouns ("I," "me," "my") in white participants, this correlation was absent in Black participants. Additionally, white people used more language to describe feelings of belongingness ("weirdo," "creep"), self-criticism ("mess," "wreck"), being an anxious-outsider ("terrified," "misunderstood"), self-deprecation ("worthless," "crap"), and despair ("begging," "hollow") as depression severity increased, but there was no such correlation for Black people. For decades, clinicians have been aware of demographic differences in how people express depressive symptoms, and this study now demonstrates how this can play out in social media.

via University of Pennsylvania and the National Institute on Drug Abuse: Sunny Rai et al, Key language markers of depression on social media depend on race, Proceedings of the National Academy of Sciences (2024). DOI: 10.1073/pnas.2319837121


First languages of North America traced back to two very different language groups from Siberia
Apr 2024, phys.org

Nichols' techniques involve the use of linguistic typology, a field that involves comparing languages and organizing them based on shared criteria. To learn more about early North American languages, she compiled lists of language characteristics and applied them to all known languages. She then scored each of the languages based on the revealed qualities. This allowed her to compare the languages as a way to find resemblances among them and spot patterns.

Nichols found that she could trace the languages spoken in early North America back to just two lineages, both of which originated in Siberia. They came, she notes, with the people who made their way across land bridges during Ice Age glaciation events.

via University of California Berkeley: Johanna Nichols, Founder effects identify languages of the earliest Americans, American Journal of Biological Anthropology (2024). DOI: 10.1002/ajpa.24923