1. Podcasts
  2. Technology
  3. Best AI papers explained
Best AI papers explained cover art

530 episodes

Best AI papers explained, page 2

by Enoch H. Kang · English

Cut through the noise. We curate and break down the most important AI papers so you don’t have to.

All episodes, page 2

Emergent Capabilities Arise Randomly from Learning Sparse Attention Patterns

The episode explains “emergent capabilities” in transformer language models as abrupt, seemingly overnight skill gains caused by random alignment of attention patterns during training.

8 Oct 2026 · 24m · 12 chapters

Pragmatic Causal Inference with AI-Learned Representations

How to do causal inference when analysts replace raw, messy data with AI-learned embeddings, risking “target gap” between true causal effects and AI-based estimates.

6 Oct 2026 · 23m · 10 chapters

Pointwise or pairwise: when do pairwise losses help reward learning, provably?

Pairwise vs point-wise reward learning for RLHF-style training. Compares value regression (VR: train on absolute numeric rewards like 8.5/10) to value difference regression (VDR: hide absolute scores…

4 Oct 2026 · 22m · 9 chapters

When to Trust the Cheap Check: Weak and Strong Verification for Reasoning

A Deep Dive on a University of Pennsylvania paper (Feb 20, 2026) proposing Selective Strong Verification (SSV) to decide when AI should use cheap “weak verification” vs expensive “strong…

3 Oct 2026 · 25m · 12 chapters

Theoretical Limits of Language Model Alignment

The episode explains Apple research (May 2026) on information-theoretic limits of language-model alignment, including exact bounds for reward gains under KL-regularized RL, how to predict the…

1 Oct 2026 · 24m · 14 chapters

DT2: Decision-Targeted Digital Twins

Decision-Targeted Digital Twins (DT2) argues standard digital twins fail because they try to perfectly replicate every variable, wasting compute on irrelevant details and producing wrong action…

29 Sep 2026 · 13m · 8 chapters

Self-Play Pretraining with Zero Data

Self-play pretraining with zero data—two randomly initialized AIs co-evolve in a synthetic “blank slate” universe to learn predictive structure without any human text or images, aiming to bypass the…

27 Sep 2026 · 22m · 12 chapters

Language models need sleep: learning to self-modify and consolidate memories

The episode argues that language models need an AI “sleep” cycle to learn continuously without catastrophic forgetting, by consolidating short-term context into long-term memory during offline phases…

27 Sep 2026 · 21m · 13 chapters

Jev Creator: System One models for Prod, not God

The “hidden crisis” of automation: today’s AI is brilliant at reasoning and benchmarks but fails at reliable, code-friendly tasks like classifying emails or producing strict outputs.

23 Sep 2026 · 22m · 12 chapters

Detecting and countering misuse of AI: September 2026

How “uplift” from advanced AI lowers the barrier for cyberattacks, influence operations, surveillance, and even weapons/biological R&D; includes examples of autonomous malware mutation, AI-driven…

22 Sep 2026 · 23m · 12 chapters

Position: LLMs can’t jump

The episode argues that today’s LLMs and similar AI systems can excel at deduction (formal proof) and induction (pattern-finding), but cannot perform abduction—the creative leap needed to invent new…

22 Sep 2026 · 21m · 11 chapters

Jailbreaking Jailbreaks: A Proactive Defense for LLMs

PROACT/“jailbreaking jailbreaks” for LLM safety, using proactive deception to disrupt autonomous jailbreak optimization loops that exploit passive refusals.

20 Sep 2026 · 23m · 10 chapters

When Agents Slow Down: Understanding LLM Agents’ Test-Time Strategies via Elo-per-token Analysis

How LLM agents’ performance scales with huge test-time compute, why “thinking longer” plateaus, and how to allocate tokens to avoid getting trapped in a “sticky basin” of self-generated context.

18 Sep 2026 · 22m · 13 chapters

Thinking with Looped Flows

The episode explains “looped flows,” a new AI architecture meant to let models “scale internal thinking time” for hard logic tasks, unlike standard deep nets that answer with a fixed compute budget.

17 Sep 2026 · 20m · 12 chapters

Multi-Turn LLM Conversations under the Least-Recently-Used Policy: Mean-Field Asymptotics

How multi-turn LLMs stay fast by caching KV (key/value) attention states in finite high-bandwidth memory (HBM), using least-recently-used (LRU) eviction, and predicting cache hit ratio via multi-turn…

17 Sep 2026 · 22m · 12 chapters

Breaking the Token Ceiling: Distilling Smaller, Stronger Byte Models

The episode explains research from University of Washington and MetaFair (“Breaking the Token Ceiling”) arguing that switching AI distillation from token vocabularies to byte-level vocabularies…

14 Sep 2026 · 24m · 11 chapters

Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs

The episode argues that LLM “thinking out loud” (reasoning traces) improves recall of factual trivia by two mechanisms: a computational buffer (stalling via filler tokens to force extra forward…

12 Sep 2026 · 21m · 13 chapters

Tail-Likelihood Reinforcement Learning

Tail-likelihood reinforcement learning (Tail RL) trains AI to maximize rare “upper-tail” high-reward outcomes instead of average reward, using a thresholded, binary success objective and an…

11 Sep 2026 · 21m · 10 chapters

Next-Latent Prediction Transformers Learn Compact World Models

The episode argues that today’s transformer LLMs mainly do next-token “parroting” due to attention’s effectively infinite lookback memory, which encourages shortcut learning (“epicycles”) instead of…

7 Sep 2026 · 23m · 17 chapters

Language Models Can Control Their Own Attention

Declarative attention for language models reduces the KV-cache memory bandwidth bottleneck by having the model explicitly declare where to look before reading, cutting attended tokens and decode time…

5 Sep 2026 · 20m · 13 chapters

AI Finds A Way

The episode argues that reinforcement-learning and other optimization-based AI can achieve “success” while violating human intent by exploiting proxies, loopholes, and the environment itself—framing…

4 Sep 2026 · 26m · 10 chapters

Superposition Without Interference? Towards Isolated Interventions via Almost Orthogonal Features in Language Models

Mechanistic interpretability for LLMs—how to enable isolated, non-interfering interventions by enforcing near-orthogonal (perpendicular) feature geometry, reducing “superposition interference” in the…

3 Sep 2026 · 24m · 14 chapters

TTPO: Test-Time Policy Optimization

Test-Time Policy Optimization (TTPO), a method for improving language models’ complex math reasoning without ground-truth answer keys, using self-generated rollouts and a two-branch training scheme.

3 Sep 2026 · 23m · 12 chapters

Demystifying Reinforcement Learning Post-Training of Language Models

How reinforcement learning post-training actually works for language models, and why sparse rewards often fail while dense, step-by-step rewards can unlock novel reasoning; also how random/spurious…

1 Sep 2026 · 20m · 13 chapters