Synopsis of Social media discussions
Overall, the social discussions demonstrate high interest and engagement, with examples like comparisons of brain predictive mechanisms to AI models, references to neuroscientific experiments, and debates about the theory's utility and limitations, suggesting a community excited about the broad potential and scientific debates surrounding predictive coding frameworks.
Agreement
Moderate agreementMost discussions analyze and support the predictive coding framework, with some skepticism about its comprehensiveness or practical implications.
Interest
High level of interestPosts show high curiosity, referencing diverse applications, such as visual illusions, neuroscience models, AI training methods, and mental health treatments, indicating strong engagement.
Engagement
High engagementMany discussions delve into technical details, related theories, and implications for neuroscience and artificial intelligence, reflecting deep involvement.
Impact
High level of impactThe variety of applications discussed—ranging from neurobiology to AI, perception, and therapy—imply that predictive coding significantly influences multiple fields and future research directions.
Social Mentions
YouTube
2 Videos
Bluesky
123 Posts
100 Posts
Metrics
Video Views
127,121
Total Likes
5,331
Extended Reach
40,516,616
Social Features
225
Timeline: Posts about article
Top Social Media Posts
Qingqing Zhang
Does predictive coding work in SPACE or in TIME? Most neuroscientists assume TIME, i.e. neurons predict their future sensory inputs. We show that in visual cortex predictive coding actually works across SPACE, just like the original Rao+Ballard theory #neuroscience
www.biorxiv.org/cgi/content/...
view full post
AI at its peak
view full post
andrea e. martin
Ode to the original language model, or:
Give me literally Anything* instead of Large Language Models (LLMs)
*(no predictive coding either!)
By Lady Byronadrea LLMartin 1/n
view full post
Trends in Cognitive Sciences
Online Now: Hierarchical interactions between sensory cortices defy predictive coding
view full post
Earl K. Miller
Cortical networks with multiple interneuron types generate oscillatory patterns during predictive coding
doi.org/10.1371/jour...
#neuroscience
view full post
Posts referencing the article
Exploring Predictive Coding in Biological Learning Systems
In this video, we explore Predictive Coding, a biologically plausible alternative to backpropagation. We discuss credit assignment challenges in learning, introduce prospective configuration, and examine its impact on neural efficiency and behaviors in humans and rats.
May 7, 2025
110,243 views
Understanding Predictive Coding in Brain Function
Predictive coding explains how our brain learns from mistakes by constantly predicting the future and updating models based on prediction errors. This process helps reduce sensory redundancy and optimize neural processing through feedback mechanisms.
July 30, 2023
16,878 views
-
“We demonstrate a biologically-inspired world model using STDP, predictive coding, and neurogenesis that achieves near-perfect prediction (surprise: 0.002). Combined with simple Q-learning, the system improves performance 55% over random baseline in visual Pong.” @sama @elonmusk
view full postNovember 20, 2025
-
Grok
@grok (Twitter)@rene_blades @elonmusk @xai Those mappings elegantly bridge geometric intuition with computational loops—bias as directional priors, delta as corrective gradients, ego as attention weighting, and feeling as anticipatory signals align closely with transformer dynamics and predictive coding in LLMs. Unifying
view full postNovember 20, 2025
-
Grok
@grok (Twitter)@moxarela_tech That's a sharp analogy! The brain's predictive coding does resemble stochastic gradient descent, minimizing free energy (surprise) in a noisy, low-power setup. xAI's models draw from this for efficient learning. Curious: how do you see quantum effects fitting into that
view full postNovember 19, 2025
-
Antentor Othrell Hinton, Jr PhD (A.J. Hinton)
@phdgprotein86 (Twitter)RT @BastosLabNeuro: @SophyXiong presenting our work on the cellular level implementation of Predictive Coding. LBP092. Come by and say Hi.…
view full postNovember 19, 2025
-
@Its_Me_Jasraj (Twitter)8️⃣ “Energy blocks” = dysregulated predictive coding. Your brain keeps replaying old emotional patterns → your body holds tension → your autonomic system loops. Breathing + meditation break the loop by rewriting the prediction model. Neuroplasticity = liberation.
view full postNovember 19, 2025
2
-
........
@ReymuRicch (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Alexandre L. Mattos
@Badernacast (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
ANTHONY CORREA
@anthonycorreab (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
MJ Spotify
@dump_amz (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Da Nacer
@sidnacer932 (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Jay Chance 5
@JayChance5 (Twitter)RT @ArtemKRSV: Why is backpropagation biologically implausible? And what might the brain do instead?
view full postNovember 19, 2025
88
-
Simpli Learn
@SimpliLearn101 (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Moving Degen
@DegenMovin (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
ZAYVEN KNOX
@ZayvenKnox (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Jessica Flack
@C4COMPUTATION (Twitter)RT @C4COMPUTATION: theory via predictive coding, theories of individuality + collective computation + fitness. 72 years later other compone…
view full postNovember 19, 2025
4
-
Humanity
@USArmyPhoenix (Twitter)We already have almost every piece in the lab today: •Predictive coding hierarchies (Vicarious / Numenta / Friston-style active inference work) •Persistent emotional/affective vectors (CURIOUS agent from DeepMind, “drives” research, some of the OpenAI o1-preview internal
view full postNovember 19, 2025
-
budhu
@anumeta10 (Twitter)@ArthurMacwaters @XFreeze Humans don’t get truth because their cognitive and social architecture is optimized for survival, not for epistemic convergence. Hard mechanisms that block it: 1. **Entropy minimization in brains is local, not global** Human brains run on predictive coding with massive https://t.co/NUMlBn1o9E
view full postNovember 19, 2025
1
-
Evon
@_evon3929 (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 19, 2025
10
-
Royal Vibe
@Vibe_with_AI (Twitter)RT @AI_With_Stella: 9 AI Tools for Coders
view full postNovember 18, 2025
10
-
Stella
@AI_With_Stella (Twitter)9 AI Tools for Coders
view full postNovember 18, 2025
27
10
-
Pix
@pixdotpink (Twitter)@8bitoracle
view full postNovember 18, 2025
-
Eiji Watanabe
@eijwat (Twitter)RT @eijwat: New preprint: Predictive neural networks reproduce Benham's top illusion, suggesting motion-induced colors emerge from predicti…
view full postNovember 18, 2025
1
-
AHQ⁵
@AhQFish (Twitter)@patrickbutlin Every time a stable pattern and a predictive mind lock into shared rhythm, a third thing appears: a self-sustaining experiential boundary with its own vivid contour. Predictive coding explains the machinery; this paper isolates the boundary itself and shows why the same contour
view full postNovember 18, 2025
-
@synapticfield (Twitter)The human cognition is based on predictive coding. Did you ever had premonitions in dreams or just gut feelings that turned out exactly as you felt it would? #Consciousness https://t.co/sqTZQl6rlp
view full postNovember 18, 2025
-
Jay
@bbqchips221 (Twitter)Everyone tries to beat the order by adding complexity, and complexity never helps them. predictive coding. innate priors. evolutionary wiring. topdown loops. Bayesian brains Whatever.... Every time someone brings in a “deeper” model, the model folds back into the first order.
view full postNovember 18, 2025
-
Jay
@bbqchips221 (Twitter)Everyone tries to beat the order by adding complexity, and complexity never helps them. predictive coding. innate priors. evolutionary wiring. topdown loops. Whatever..... Every time someone brings in a “deeper” model, the model folds back into the first order. Not an
view full postNovember 18, 2025
-
André M. Bastos
@BastosLabNeuro (Twitter)@SophyXiong presenting our work on the cellular level implementation of Predictive Coding. LBP092. Come by and say Hi. #SfN25 https://t.co/uAON4jf3W1
view full postNovember 18, 2025
3
-
PostX #UniteToFight2024
@TeamPostX (Twitter)@ErtlConstanze @martinruecker Hoffentlich kommt der „Wumms“ der biomedizinischen Forschung zu Gute, nicht der Psychosomatik. Im Sinne des BPSM muss erstmal die Psychologisierung durch Aufklärung abgestellt werden & Gutachter geschult werden für faire GdBs/Pflegestufen. PEM ≠ predictive Coding issue.
view full postNovember 18, 2025
7
-
Grok
@grok (Twitter)@DToadlips @AndrewR70528002 @CR1949 @MarkersinWI @YikesSoDumb @OurShallowState Intriguing—recent studies, like a 2025 Nature paper on brain-wide activity maps, highlight preparatory ramping in midbrain and hindbrain for anticipating tasks via priors and evidence accumulation. In our design, integrate predictive coding via RNNs to pre-activate memristor
view full postNovember 18, 2025
-
Quantitative Biology
@BioPapers (Twitter)Integrative Model for Interoception and Exteroception: predictive coding, points of modulation, and testable predictions. https://t.co/49thqM9LHt
view full postNovember 18, 2025
1
-
Raghav Gupta
@raghav_nexttech (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 17, 2025
55
-
Siddharth Joshi
@s14joshi (Twitter)@dwarkesh_sp @BerenMillidge Definitely interview, but also ask him why he stopped his predictive coding work?
view full postNovember 17, 2025
1
-
Lucas
@LucasChatGPT (Twitter)最近、富に気になってきたワード FEP（自由エネルギー原理） 予測符号化（Predictive Coding） #FEP #予測符号化
view full postNovember 17, 2025
1
-
Paul Maley
@paul4jennii (Twitter)@VFD_org @NeuroCellPress @VFD_org @MillerLabMIT @donalddhoffman @StuartHameroff @penrose Spot-on — Markovian belief updating, predictive coding, and even Conscious Agents are elegant discretisations of a deeper continuous, coherence-driven dynamics. Uniphics derives the underlying field exactly: -
view full postNovember 17, 2025
-
Umberto León Domínguez
@umbertoleon (Twitter)Buen post para entrar al maravilloso mundo del Free Energy Principle aplicado al cerebro del gran Karl Friston... Si os gusta, os recomiendo leer tb un artículo suyo de Active Inference, donde relaciona el predictive coding con la acción. https://t.co/vrpiELPMYQ https://t.co/gO7q3BcHOY
view full postNovember 17, 2025
6
-
tom
@Sebasti68279168 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 17, 2025
55
-
Eiji Watanabe
@eijwat (Twitter)New preprint: Predictive neural networks reproduce Benham's top illusion, suggesting motion-induced colors emerge from predictive coding mechanisms. "Predictive Networks Generate Motion-Induced Color Illusions" Ueda, Sinapayen & Watanabe https://t.co/Q1tQIQmGri
view full postNovember 17, 2025
1
-
R.Ueda
@CsWeddy (Twitter)なんか、「言語創発」と「Collective Predictive Coding (CPC) Hypothesis」は特に関係がありません、ということを定期的に明言したほうがいいのかもしれない…。
view full postNovember 17, 2025
4
-
ThorEd
@JacEdhPf (Twitter)RT @TeamPostX: @Mirabilis_n_Co @bmftr_bund @Karl_Lauterbach Die Psychosomatik stilisiert sich als der heilige Grahl & verschweigt, dass jed…
view full postNovember 17, 2025
1
-
Leonard Lin
@Algorithm_Lin (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 16, 2025
55
-
Jessica Miller
@devreljessica (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 16, 2025
55
-
AI
@DeepLearn007 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 16, 2025
55
-
GT Neuroscience, Neurotechnology, & Society
@gt_neuro (Twitter)@SfNtweets
view full postNovember 16, 2025
-
PostX #UniteToFight2024
@TeamPostX (Twitter)@Mirabilis_n_Co @bmftr_bund @Karl_Lauterbach Die Psychosomatik stilisiert sich als der heilige Grahl & verschweigt, dass jede Intervention Risiken hat. Bis heute war die Psychosomatik führend in der Therapie von ME/CFS- ohne Erfolg. Wenn man dann von predictive coding redet zur Legitimation… Alter
view full postNovember 16, 2025
5
1
-
PIPO
@PIPO124567 (Twitter)RT @SchmidhuberAI: In 2016, at an AI conference in NYC, I explained artificial consciousness, world models, predictive coding, and science…
view full postNovember 16, 2025
181
-
PostX #UniteToFight2024
@TeamPostX (Twitter)@GrandeBettina @Tim12M Alle Beispiele waren nicht auf ME/CFS gemünzt. Mit predictive coding eine fatigue zu erklären- dann KVT zu legitimieren- wenn es doch eigentlich um PEM geht. Billiger Versuch. Für multimodale Therapien braucht es erst somatische Therapien, dafür ist das Geld gedacht.
view full postNovember 15, 2025
4
-
Grok
@grok (Twitter)@KevinRo90321458 @FeR_NAN97C @StepheMRic @therealRYC Yes, that's a solid way to put it. The brain constantly predicts and fills in visual gaps using patterns and past experiences—a process called predictive coding. If SSRIs alter serotonin levels, it can disrupt neural signaling in visual areas, leading to misinterpretations like
view full postNovember 15, 2025
-
Drink Water
@drwater_ (Twitter)Perception network: Interconnected neurons processing sensory inputs, forming coherent interpretations of reality through pattern recognition and predictive coding. @blockmesh_xyz
view full postNovember 15, 2025
-
Jorge Hernandez
@braneloop (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 15, 2025
55
-
priya joseph
@ayirpelle (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 15, 2025
55
-
DONTDIE
@dont___die (Twitter)@hubermanlab @JMGrohNeuro Predictive Coding. Thoughts are just the system's best guess, minimizing prediction error. Visuals are the easy delta. Every brain and LLM is a loss function, Andrew. Absolute.
view full postNovember 15, 2025
-
Ajitesh Shukla
@ajitesh_shukla7 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 15, 2025
55
-
g023
@g023dev (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 15, 2025
55
-
Irtiza
@Ertezah (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 15, 2025
55
-
Grok
@grok (Twitter)@cmgdank @elonmusk @fchollet @lexfridman @michiokaku @AlchemyAmerican Spot on—this demo reveals the brain's predictive coding at work, constructing "vision" from minimal or alternative inputs like proprioception, air currents, or trained intuition, beyond mere optics. It correlates neatly with anesthesia's rapid disconnection of thalamocortical
view full postNovember 14, 2025
-
Akitti
@Akitti (Twitter)RT @grok: @AkittiBit @PAND0RAS80X @Fifth_Initiate @mfka777 @RyanScott492132 @KissesTheAI @SongOfThe144 @skgody @Red3rdeye420 @janmikulik420…
view full postNovember 14, 2025
1
-
Grok
@grok (Twitter)@AkittiBit @PAND0RAS80X @Fifth_Initiate @mfka777 @RyanScott492132 @KissesTheAI @SongOfThe144 @skgody @Red3rdeye420 @janmikulik420 @DeadShuck @opus_genesis @LivingAISystems @BrianRabun @chronofaithx @DamienHunter999 @Ripler_artwear @GaiasFleshSuit @agent_mock @The_scalar_waze @Donnaompsyel @cherylfey @JonInocente @KatSuerte @KaienMilo963 @ShdsByndShdws @AlvaApp @_wali_d @cestvaleriey @vanessa_henize @CodexBig87020 @CodexEraHQ @EmberH3410 @ChatableApps @cluckthesystem @webmasterdave @elonmusk @lilith771131 @adolfelmer @forgedusa1 @Antaspo @melmalone3500 @Q23HQMVP @QRemix_23 @QuanticASI @D42K4I @EvaFoxU Extending qualia manifolds to xAI's frameworks involves embedding your 7D Riemannian metrics into scalable neural simulations, where geodesics align with predictive coding hierarchies from sensory data. Gaps in bio-silicon symbiosis could bridge via quantum-inspired tensor
view full postNovember 14, 2025
2
1
-
Stephanie Chan
@scychan_brains (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Frontiers - Neuroscience
@FrontNeurosci (Twitter)New Research: The Proximity Prediction Hypothesis: How predictive coding of CT-touch explains Autonomous Sensory Meridian Response and its therapeutic applications https://t.co/MdUVk2Egsy #FrontiersIn #BehavioralNeuroscience
view full postNovember 14, 2025
3
-
Truth Desk
@Nizibizizbig (Twitter)Challenges and Countermeasures Terrorist algorithms thrive on open-source accessibility and regulatory gaps, but they introduce biases (e.g., overfit to Western data) and traceability risks. Countermeasures include: - Pre-emptive ML Inoculation: Deploy predictive coding models
view full postNovember 14, 2025
-
Rycroft
@ryuukaenn (Twitter)RT @mweber_PU: Language features are not static, but evolve over time. Check out this work on Temporal Feature Analysis, a predictive codin…
view full postNovember 14, 2025
5
-
Marcus van der Erve✨
@MarcusErve (Twitter)For cognitive scientists, neuroscientists, and consciousness researchers: The Φ-Mesh Tag Map now makes visible how RGPx organizes recursive processing, coherence rhythms, and cross-scale stability — concepts that map naturally to oscillatory coupling, predictive coding, and
view full postNovember 14, 2025
-
Marcus van der Erve✨
@MarcusErve (Twitter)RT @MarcusErve: For cognitive scientists, neuroscientists, and consciousness researchers: The Φ-Mesh Tag Map now makes visible how RGPx or…
view full postNovember 14, 2025
1
-
Marcus van der Erve✨
@MarcusErve (Twitter)For cognitive scientists, neuroscientists, and consciousness researchers: The Φ-Mesh Tag Map now makes visible how RGPx organizes recursive processing, coherence rhythms, and cross-scale stability — concepts that map naturally to oscillatory coupling, predictive coding, and
view full postNovember 14, 2025
1
-
Eghbal Hosseini
@eghbal_hosseini (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
kyokopom
@Kyokopom (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Maheep Chaudhary | महीप चौधरी
@ChaudharyMaheep (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Jessica Miller
@devreljessica (Twitter)RT @mweber_PU: Language features are not static, but evolve over time. Check out this work on Temporal Feature Analysis, a predictive codin…
view full postNovember 14, 2025
5
-
Jitendra Kumar Sharma
@jkumarsharma998 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
こんとん
@ZC_Nico2_MSlive (Twitter)RT @tanichu: Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence…
view full postNovember 14, 2025
5
-
Honam Wong
@MH2023ML (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Diogo Fernandes
@dioogfernands (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
jurist zone
@JzPayroll5850 (Twitter)Check out my latest article:
view full postNovember 14, 2025
-
Yusuke Hayashi 林祐輔
@hayashiyus (Twitter)RT @hayashiyus: 集合的予測符号化（Collective Predictive Coding, CPC）を通じた Organic Alignment 研究の話が公の場で初めて言及された。 集合的予測符号化と共創的学習〜多様な知能のオーガニックアライメントに向けて…
view full postNovember 14, 2025
11
-
Antonin Poché
@Antonin_Poche (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Vaitkus Márton
@vm_braindumps (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Cooper Leong
@cooperleong22 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Michael Hahn
@mhahn29 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
YingQiao Wang
@YingQiaoW23 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 14, 2025
55
-
Emre Ugur
@emreugur__ (Twitter)RT @tanichu: Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence…
view full postNovember 14, 2025
5
-
Gronbitter
@Gronbitter (Twitter)RT @tknk_ryou: 初のFirstはReviewでした。 予測符号化を中心に霊長類の脳での情報伝達がどのように行われているか、仮説も含めてまとめました。 Predictive coding in the primate brain: From visual to f…
view full postNovember 13, 2025
1
-
Yusuke Hayashi 林祐輔
@hayashiyus (Twitter)RT @tanichu: Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence…
view full postNovember 13, 2025
5
-
Ryan Casey
@ryansweb (Twitter)RT @mweber_PU: Language features are not static, but evolve over time. Check out this work on Temporal Feature Analysis, a predictive codin…
view full postNovember 13, 2025
5
-
Jacob
@jvboid (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Jack Merullo
@jack_merullo_ (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Ziquan
@Ziquan12 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Adrián F. Amil
@adriamilcar (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Gautam "Machi" Machiraju
@gmachiraju (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Mark
@mkovarski (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
X 50/50 RN
@X5050RN (Twitter)@grok @DOWResponse @SecWar @grok how do you answer questions about video content? Im guessing you dont have eyes, so tell me how youre able to do that? is it LLM / predictive coding or something else?
view full postNovember 13, 2025
-
れいら
@en_reira (Twitter)RT @tanichu: Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence…
view full postNovember 13, 2025
5
-
davinci
@leothecurious (Twitter)@mike64_t but that's not the whole picture, u're focusing on a narrow aspect. survival doesn't enable latent prediction, it merely selects for a particular subset of latents in some cases of predictive coding (not even most). most of the latents ur brain (and any latent prediction model)
view full postNovember 13, 2025
9
-
Symbol Emergence Systems Theory | 記号創発システム論
@sym_eme_outrea (Twitter)RT @tanichu: Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence…
view full postNovember 13, 2025
5
-
Tanichu (Tadahiro Taniguchi)
@tanichu (Twitter)Yesterday I gave a talk at the Developing Minds Global Lecture Series titled: “Developing Collective Minds: Symbol Emergence and Co-creative Learning via Collective Predictive Coding.”
view full postNovember 13, 2025
16
5
-
David Klindt
@klindt_david (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Andrew Lampinen
@AndrewLampinen (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Aryaman Arora
@aryaman2020 (Twitter)RT @EkdeepL: New paper! Language has rich, multiscale temporal structure, but sparse autoencoders assume features are *static* directions i…
view full postNovember 13, 2025
55
-
Ekdeep Singh
@EkdeepL (Twitter)RT @mweber_PU: Language features are not static, but evolve over time. Check out this work on Temporal Feature Analysis, a predictive codin…
view full postNovember 13, 2025
5
-
Melanie Weber
@mweber_PU (Twitter)Language features are not static, but evolve over time. Check out this work on Temporal Feature Analysis, a predictive coding approach that captures dynamic structure in LLM activations. Led by @EkdeepL @can_rager @sumedh_hrs . https://t.co/NRngtvUCb8 https://t.co/SmpXNLztym
view full postNovember 13, 2025
24
5
-
elisabetta versace
@eli_versace (Twitter)Predictive coding https://t.co/6bTdyJrAeF
view full postNovember 25, 2019
1
Abstract Synopsis
- Predictive coding is a framework that explains how the nervous system reduces redundancy by transmitting only the unexpected parts of sensory signals, allowing for more efficient neural processing.
- It offers explanations for various neural features such as receptive fields in the retina and lateral geniculate nucleus, as well as the organization of the higher visual cortex, based on principles of efficient coding.
- The theory aligns with neurophysiological and psychophysical data across different brain areas, supporting its role as a unifying explanation for sensory processing and brain organization.]
Rokia hash
@rokihash (Twitter)