#computational neuroscience

[louie-and-lace]

thesis so bad it deformed my white blood cells.

[nikjag]

hashtag my substack

Nymira's Substack | Substack

My personal Substack. Click to read Nymira's Substack, by Nymira Nishita, a Substack publication. Launched 6 days ago.

[kamalkafir-blog]

Longer scans boost prediction and cut costs in brain-wide blockociation studies

Datasets, phenotypes and participants Following previous studies, we considered 58 HCP phenotypes59,60 and 36 ABCD phenotypes15,39. We also consider a cognition factor score derived from all phenotypes from each dataset31, yielding a total of 59 HCP and 37 ABCD phenotypes (Supplementary Table 4). In this study, we used resting-state fMRI from the HCP WU-Minn S1200 release. We filtered…

[in-sightpublishing]

Conversation with Dr. Marc-Oliver Gewaltig on Practical Applications of AI, Instead

Scott Douglas Jacobsen In-Sight Publishing, Fort Langley, British Columbia, Canada Correspondence: Scott Douglas Jacobsen (Email: [email protected]) Received: January 21, 2025 Accepted: N/A Published: February 15, 2025 Abstract Dr. Marc-Oliver Gewaltig, a distinguished researcher in artificial intelligence, computational neuroscience, and robotics, discusses his journey from pioneering…

Conversation with Dr. Marc-Oliver Gewaltig on Practical Applications of AI, Instead

[philosophiesde]

Zoomposium with Dr. Patrick Krauß: "Construction manual Artificial Consciousness"

In another episode of our "Zoomposium series" on the topic of "AI research" or the possibilities of a construction manual for an Artificial Consciousness (AC/DC = "artificial consciousness/digital consciousness"), my colleague Axel Stöcker from the "Blog of Big Questions" and I had the opportunity to conduct an interview with the well-known and renowned German physicist, neuroscientist and linguist Dr. Patrick Krauß.

However, for a better understanding of the interview with Dr. Krauß, here is a brief summary of the scientific articles first. Dr. Patrick Krauß and Prof. Andreas Andreas Maier (Professor of Computer Science and Head of the Department of Pattern Recognition) describe in their article "Will We Ever Have Conscious Machines?" (December 22, 2020), what consciousness could ever be in the first place from a philosophical and neuroscientific perspective according to the current state of knowledge. In the following, they try to develop their own "Construction Manual for Artificial Consciousness" mainly based on Antonio Damasio's "Theory of Consciousness".

Damasio's theory of consciousness, which he first developed in his book "The Feeling of What Happens: Body and Emotion in the Making of Consciousness" (1999), is essentially based on the postulation of 3 stages for the development of a consciousness: 1. "fundamental protoself", 2. "core consciousness" and 3. "extended consciousness". His theory for the development of a form of consciousness builds here substantially on a concept of "feeling", as the title of his book already reveals. Krauß and Maier call this consequently in modification of the Cartesian saying: "Sentio ergo sum". These new approaches presuppose, of course, a concept of "embodiment" and "embededness" in the sense of the "4E theory" of the "philosophy of situated cognition" (PSK), which I have also frequently tried to point out in my essays, e.g. "The System Needs New Structures" (https://philosophies.de/index.php/2021/08/14/das-system-braucht-neue-strukturen/). Therefore, I was all the more pleased that these "new structures" could finally find a concrete implementation in AI research for the development of Artificial Consciousness, as proposed by e.g. Krauß and Maier in their scientific papers mentioned above.

In the aforementioned article "Will We Ever Have Conscious Machines?" Krauß and Maier now try to transfer the various stages of Damasio's theory of consciousness 1:1 into concrete schematics for deep learning on machines. For this purpose, different strategies of "feed-forward connections", "recurrent connections" in the form of "reinforcement learning" and "unsupervised learning" are applied to simulate the biological processes of neuronal networks. And this is exactly what we once asked Dr. Krauß about in our Zoomposium interview. More at: https://philosophies.de/index.php/2023/10/24/bauanleitung-kuenstliches-bewusstsein/

Or: https://youtu.be/rXamzyoggCo

[findersfeeders]

We just interviewed an electrical engineering hire foisted upon us by the college dean. I suspect he wants my dept to pay the other 50% of his salary.

Dude is….not our vibe. My gos the teeth that were bared at his job talk, phew! Homedude could not cite a place he could improve. A textbook narcissist fr. His research is prolific and productive and he comes with hella connections our students need.

I asked him how his research, tuned to thalamus models of human vision, adapts to active sensing. He presents his work as though the sensors are sedentary and not strategically placed and varied across the compsite scene. He then tried to go tit for tat with me on LGN processing and organization. He was like there are channels in LGN and I’m like yes but they are no longer retinotipic, so you are building the field composition at that point. He was wrong but spoke with so much confidence that I got flustered and had to google LGB processing. I was definitely correct.

He then tried to tell me we don’t ID eyes til after V4. Like bro what??? Eyes are Gabriel functions, we’re tuning to them saccadically with info from MGN! Way before v1 processing! He said the percept of visual quality is instantaneous. Like…my guy it is not. The general rule of brain processing is the harder the task the longer the process. Information varies across the composite visual scene. Where aberrations are in the visual field affects detection. He also was like we don’t know why GSM is the underlying structure of visual processing. Hug?? Because naturalistic stimuli of all kinds follows the power law which is the FFT of a Gaussian and Gaussians dictates most cosmic phenomena. This is first year (computational) neuroscience. I don’t… I can’t. An electrical engineer stumbled across neuroscience and uses it to be flashy but is not actually informed. Barf on my shoes.

If I had time and funding sometime in his method and be like “you don’t have tot use the whole image just where places we are likely to look, which are driven by high contrast edges and peripheral motion.”

I think if he’s what it takes to appear the dean we (meaning I) can manage. I’m walking into the office and the ick level is rising. The childlike way he spoke to me and the implied disruption he expected to face in his classrooms and the “defensiveness” he “worked through” when bringing a pitch to the Deaf and HoH community all make me worried for the mental wellness of our students.

I’m gonna reach out to some idiot processing folks and see if they be heard of him and chat about sparse sensing. Just to check my sanity. Anyway..that’s my vent for the day.

[noosphe-re]

Discrete topological spaces and place field maps (Babichev, Andrey & Dabaghian, Yuri. (2017). Topological Schemas of Memory Spaces. Frontiers in Computational Neuroscience. 12. 10.3389/fncom.2018.00027.)

[pavel-nosok]

NASA’s Mars Rovers Could Inspire a More Ethical Future for AI

Since ChatGPT’s release in late 2022, many news outlets have reported on the ethical threats posed by artificial intelligence. Tech pundits have issued warnings of killer robots bent on human extinction, while the World Economic Forum predicted that machines will take away jobs. Read more…

View On WordPress