The hottest Science Substack posts right now

And their main takeaways
Category
Top Science Topics

A critique of successor representations as a model of learning in the hippocampus

Artificial General Ideas • 1 implied HN point • 14 Sep 24
🔬 Science Neuroscience Psychology Cognition Learning Models
  1. Successor representations (SR) does not explain how place cells in the hippocampus learn or form. It assumes inputs that are already perfect place fields, so it can't help in understanding their development.
  2. Many claims about SR's abilities, like making predictions or forming hierarchies, actually relate to simpler models like Markov chains. SR doesn't add much value to those features.
  3. Experiments often used to support SR in humans might actually show evidence for more general planning methods. Model-based reasoning seems to fit the observed behavior better than SR does.

💸 Living Ink unveils details of $3.5M funding to scale carbon-negative black pigment

paxtier • 2 implied HN points • 12 Dec 23
🔬 Science Aquaculture Funding Market Trends Research Climate action
  1. Living Ink received $3.5M funding for carbon-negative black pigment scaling
  2. Challenges in offshore low-trophic aquaculture include messaging, business cases, and long-term funding
  3. Global markets see significant developments in sustainable agriculture, blue carbon initiatives, and innovative algae products

Halting prophecy

Notices to three friends • 2 implied HN points • 01 Dec 23
🔬 Science Computer Science Prediction Self-understanding Limitations
  1. The halting problem in computer science cannot be solved because of the limitations of predicting a program's behavior.
  2. Prophecy and prediction face conceptual limitations due to the inability to fully control or predict the future.
  3. The connection between the halting problem and prophecy reveals insights about self-understanding, unpredictability, and the quest for knowledge.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Space is a sensory-motor sequence in the hippocampus

Artificial General Ideas • 1 implied HN point • 12 Aug 24
🔬 Science Neuroscience Cognitive Science Psychology Biology Artificial Intelligence
  1. The hippocampus may not just represent physical space but instead processes space as a sequence of sensory and motor experiences. This means how we perceive space comes from our interactions, not just where we are.
  2. Place cells in the brain react to specific sequences of observations rather than directly to locations themselves. This explains why experiences in different environments can create similar neural responses.
  3. New models, like causal graphs, allow for better understanding and planning in navigational tasks. They can adapt to new environments quickly by using learned sequences without needing to rely on exact spatial representations.

Axial Discovery - Phenotypic screening

Discovery by Axial • 3 implied HN points • 27 Mar 23
🔬 Science Drug Discovery Genomics Drug Development Biomedical Research
  1. Phenotypic screening focuses on identifying specific physical or biochemical traits of interest for drug discovery.
  2. Key rules for effective phenotypic screens include selecting relevant cell models, designing disease-specific assays, and defining clinical-like endpoints.
  3. Advancing phenotypic screening requires improving throughput of complex models, developing translational disease models, enhancing proteomic tools, and integrating phenotypic and target-based screening.

Epsilons, no. 3: The LU decomposition

The Palindrome • 3 implied HN points • 27 Mar 23
🔬 Science Mathematics Linear Algebra
  1. Matrix factorizations are a key part of linear algebra, used for inverting matrices and simplifying determinants.
  2. The LU decomposition method involves breaking a matrix into upper and lower triangular forms.
  3. Linear algebra helps in solving systems of linear equations by transforming them into echelon form using operations like multiplying by scalars and adding equations.

Axial Discovery - Machine learning and DNA-encoded libraries

Discovery by Axial • 3 implied HN points • 13 Feb 23
🔬 Science Machine Learning
  1. DNA-encoded libraries (DEL) use unique DNA barcodes to screen billions of chemical compounds efficiently.
  2. Machine learning is being utilized in DELs to train models for virtual screening and map out structure-activity relationships more rapidly.
  3. Challenges in DELs include improving diversity, developing better filters for virtual screening, and expanding screens to select for features like toxicity and ADME.

DiffGrad : Is it the right optimization method for training your CNNs?

Machine Learning Diaries • 2 HN points • 25 Sep 23
🔬 Science Optimization Machine Learning Neural Networks Research
  1. Optimizing neural networks with DiffGrad may prevent slow learning and jittering effects in training
  2. DiffGrad adjusts learning rates based on gradient behavior for each parameter, leading to improved optimization
  3. Comparisons suggest that DiffGrad outperformed Adam optimizer in terms of avoiding overshooting global minima

Is the “Receptor for AGEs (RAGE)” Really a Receptor for AGEs?

Harnessing the Power of Nutrients • 39 implied HN points • 21 Nov 10
🔬 Science Research Health Inflammation Proteins
  1. The Receptor for AGEs (RAGE) may not actually bind to and activate AGEs as previously believed, leading to a debate in the scientific community
  2. Studies have shown that RAGE is involved in processes like cell proliferation, tissue repair, and immune system function, rather than solely causing inflammation
  3. Deletions of the RAGE gene in mice have shown varied effects, indicating the complex role RAGE plays in the body

Biological Age

Adjacent Possible • 2 HN points • 13 Jun 23
🔬 Science Life Extension
  1. Part Three of Immortality explores the reasons behind aging and the concept of death as a life-changing force.
  2. Studies on extending the lives of worms could challenge our understanding of human aging and open up possibilities for life extension.
  3. Consider subscribing for a 7-day free trial to read more and access full post archives on Adjacent Possible.

Uploads are Impossible

PashaNomics • 2 implied HN points • 12 May 23
🔬 Science Physics Computer Science Philosophy Culture AGI
  1. Creating digital uploads of human minds is likely impossible due to challenges in physics, computer science, and philosophy.
  2. The process of verifying a successful upload is complex, involving difficult tasks such as identifying 'soul' in the digital mind.
  3. Cultural dynamics and human nature present challenges in ensuring the safety and ethical treatment of digital uploads.

Biological DOOM

De Novo • 2 HN points • 29 Apr 23
🔬 Science
  1. DOOM, a classic 1993 game, ran on low computing power and simple code, making it hackable to run on unusual devices.
  2. Biological systems like cells can potentially be used for computation, utilizing methods such as nucleic acid hybridization, transcription and translation, and protein phosphorylation.
  3. Running complex games like DOOM on biological systems faces challenges in speed, memory, and system integration, with neurons and protein logic being the most promising options.

🌊 200% increase in blue economy investments

paxtier • 1 implied HN point • 18 Mar 24
🔬 Science Investments Aquaculture Biotechnology
  1. There has been a 200% increase in investments in the blue economy between 2018 and 2022, signifying a growing interest and momentum.
  2. Challenges in the sector include a massive global unmet financing gap, concentration of investments at market launch phase, and investor hesitancy.
  3. Recent developments highlight collaborations in seaweed cultivation, funding raises for startups like KAS, and advancements in sustainable aquafeeds.

Snakes and Singularities

Adjacent Possible • 1 HN point • 13 Mar 24
🔬 Science Evolution Cultural Evolution Biology Ecology Technology
  1. Snakes have evolved diverse traits through sudden phase shifts rather than incremental changes, expanding their adaptive possibilities significantly.
  2. Cultural and technological evolution also experience periods of sudden expansion and contraction of possibilities, akin to biological evolution's singularities.
  3. In addition to platform shifts, singularities in evolution can be driven by the emergence of single new traits leading to significant consequences in exploring new ecological niches.

Are we so good to simulate?

world spirit sock stack • 1 implied HN point • 04 Mar 24
🔬 Science Simulation Existence Technology Computing Resources
  1. The Simulation Argument suggests that if technologically advanced civilizations are likely to create 'ancestor simulations,' then it's probable we are currently living in one.
  2. A counterargument questions the high cost and resources needed to run simulations of confused minds, suggesting that the majority of minds in a location are likely correct about their reality.
  3. The idea that simulating history is extremely cheap challenges the assumption that all possibilities will be pursued given finite resources and many potential simulations.

Russia’s Nesting Satellite Problems

Ill-Defined Space • 1 HN point • 29 Feb 24
🔬 Science Space Satellites Electronics Funding Technology
  1. Russia's space industry faces challenges in meeting satellite production goals due to financial losses, lack of imported electronics, unclear strategies, resource obstacles, and conflicting priorities from top officials.
  2. Russian satellite manufacturers struggle with access to modern electronics, impacting their ability to scale up production and meet industry goals, possibly leading to prioritization of military satellite programs over commercial or civil ones.
  3. Plans for conveyor production to increase satellite output face obstacles such as scarcity of space-rated electronics, conflicting objectives between government wishes for a space station and the need for smallsat production expansion, and funding issues.

Symbolic Mereological Expression Of A New Theory On Language

Autodidact Obsessions • 1 HN point • 10 Feb 24
🔬 Science Language Logic Symbolism Philosophy
  1. Language exists in a quantum state of potential meanings until words are used, altering their semantic intentions.
  2. The application of words changes language from potential to actual, solidifying their meanings into rigid semantic intentions.
  3. A mereological framework views language as a whole composed of parts (words) and considers the relationships between these parts both before and after their usage.

It’s the Energy, Stupid

Material World • 2 HN points • 22 Feb 23
🔬 Science Energy Economy Agriculture Environment Technology
  1. Energy is a critical driving force behind the economy, impacting everything from the production of goods to the growth of vegetables like tomatoes.
  2. The availability and cost of energy play a significant role in various industries, such as greenhouse farming for vegetables and manufacturing sectors.
  3. Energy efficiency is improving, but energy remains a crucial input for processes like producing solar panels and other advanced technologies, highlighting its importance in future advancements.

Adam Grant vs. Coleman Hughes, Part 2: Causation Does Not Imply. . . Anything

Singal-Minded • 1 HN point • 31 Jan 24
🔬 Science Psychology Research Replication Causation Social Dynamics
  1. Be cautious of over-extrapolating findings from laboratory experiments to real-world situations
  2. Research in social psychology, especially in politically charged topics like race, may lack replicability and generalizability
  3. Accuracy in summarizing and citing research in public discussions and writings is important for maintaining credibility