Tootfinder

Since it was relevant to a discussion I just had on here and is something most people probably haven't thought about much (unless you've taken one of a handful of philosophy classes), I thought I'd try to lay out a key piece of Descartes' Meditations (#philosophy

Q&A with Jonah Peretti on why he sold BuzzFeed, the structure of the deal with Byron Allen, his new role as president of BuzzFeed AI, HuffPost, and more (Nilay Patel/The Verge)
https://www.theverge.com/podcast/932154/peretti-allen-buzzfeed-ai-slop-soci…

Exclusive: Jonah Peretti explains why he sold BuzzFeed
The soon-to-be-former CEO reflects on the rise and fall of BuzzFeed, its recent sale to Byron Allen, and how a pivot to AI can give the brand new relevance.

Retrieval-Augmented Generation (RAG) systems rely on pre-chunked documents, tying retrieval to arbitrary boundaries. Carles Onielfa is joining us at #bbuzz26 to explore an experimental approach that surfaces semantically relevant text spans, without chunking.
Learn more: https://2026.berlinbuzzwords.de/session/towards-chunk-less-rag/

🚨Job Alert ‼️🚨 Actually, three (3) openings:
1) Adjunct Prof with industry experience in economic #geology and ore systems science
2) Lecturer, Sr Lecturer, or Assoc Prof, hydrogeology, geometallurgy, etc
3) Computational #geophysics

Can someone explain to me how the argument “ #LLMs only produce stochastically reasonable output ” is relevant or convincing?
Like… how do you think a human brain forms a sentence? Is this a straight-forward deterministic process? Does it matter?
I use tools and processes to verify my output and so does a coding agent.
I just don't get how this is an argument that would…

Introducing our new GÉANT Association Strategy 2026-2030: Powering Knowledge for Europe 🇪🇺
Built around three core goals: strengthening human networks, creating scalable digital services, and building financial sustainability.
And at its heart is a simple mission, to stay deeply relevant for our members and for Europe.
To mark this moment, we've also refreshed our visual identity — new logo, new website.
🔗 Read the full strategy and explore the new GÉANT:

The spectral picture of self-similar collapse in the Constantin-Lax-Majda equation
Jie Xu
https://arxiv.org/abs/2607.19762 https://arxiv.org/pdf/2607.19762 https://arxiv.org/html/2607.19762
arXiv:2607.19762v1 Announce Type: new
Abstract: We give a spectral description of the self-similar collapse profile of the Constantin-Lax-Majda (CLM) equation, the $a=0$ anchor of the generalized family $w_t a\,u\,w_x = u_x\,w$, $u_x = Hw$. Linearizing about the exact profile $\Omega(y) = -y/(y^2 1/4)$ and realizing $L_0$ as a closed operator on the origin-$H^2$ space, we prove three things at $a=0$. Its essential spectrum meets the closed half-plane $\{\mathrm{Re}\,\lambda \ge -1/2\}$ in the single vertical line $\{\mathrm{Re}\,\lambda = -1/2\}$: the line is placed by a log-widening Weyl sequence, and an explicit Hardy-Mellin resolvent bound constructively empties the rest of the half-plane apart from $0$ and $1$. Its full point spectrum over $\mathbb{C}$, on the odd realization, is exactly $\{0,1\}$, the scaling and time-shift symmetry modes, with no embedded eigenvalues; removing these by the standard modulation leaves a spectral gap of $1/2$ on $X$. The linear semigroup and its exact decay rate $e^{-\tau/2}$ are computed in closed form, but on a weighted space of the conjugated variable reached from $X$ by a bounded transfer map; we keep the two separate, since $L_0$ is non-normal and a spectral gap does not by itself give a decay rate in the $X$ norm. A realization dichotomy identifies the in-strip smear of generic discretizations as the faithful spectrum of the maximal $L^2$ realization, which origin-$H^2$ removes. For $a>0$ we prove a conditional two-line inclusion for each admissible smooth focusing profile, recompute the branch $c_l(a)$ of Lushnikov, Silantyev, and Siegel as a cross-check, and record the formal scaling-relevance exponent $s^*(a) = 1/c_l(a)$, below which fractional dissipation is asymptotically subdominant in self-similar variables for fixed sufficiently regular data. The contribution is the realization-dependent spectral picture of the collapse profile itself.
toXiv_bot_toot

Build-Authorized Evidence for Opaque Calls: A Fail-Closed Rewrite-Authority Boundary
Zhonghua Yi (Toka Language Research Group)
https://arxiv.org/abs/2607.18949 https://arxiv.org/pdf/2607.18949 https://arxiv.org/html/2607.18949
arXiv:2607.18949v1 Announce Type: new
Abstract: Detached semantic facts about opaque native providers do not by themselves justify compiler rewrites: rewrite authority must be confined to the accepted fact, selected provider and build, caller, callback environment, observation, and runtime target. We present a build-authorized path-effect interface that enforces this boundary through fail-closed authorization and link receipts. The design separates receipt closure, callback-environment closure, and projection identity, and passes accepted facts to LLVM through a narrow internal API. We use one-hop topology-load reuse as a minimal observable witness of authority, not as the optimization target.
A conservative LLVM consumer reuses a pointer observation only from a noalias root or one constant nonzero projection. Rocq models prove conditional refinement and authority non-amplification under explicit effect, alias, compiler/ABI, and target-resolution premises. We instantiate checked production with Toka: a source-summary gate emits exact LLVM IR, a separate IR checker accepts only a bounded topology-preserving subset, and only accepted IR is compiled into the receipt-bound provider object. A bounded static Darwin/arm64 profile also checks the final direct branch target.
Across issuer-declared readv, recvmsg, and Cairo boundaries, authorized IR retains each opaque call, reduces the relevant loads from two to one, and preserves observed results; mismatched providers, builds, callbacks, projections, and unsupported IR remain neutral. A libjpeg case is rejected because its callback environment is open, while a bound callback singleton demonstrates the supported closure rule. The contribution is a checked deployment-compiler boundary with an explicit trust and applicability frontier, not a uniquely expressive effect encoding or a new load-elimination algorithm.
toXiv_bot_toot

Eccentricity-Constrained CNN Training Reveals Adaptive Information Coding Around the Visual Field
Dylan M. Diaz, Margaret M. Henderson
https://arxiv.org/abs/2607.19316 https://arxiv.org/pdf/2607.19316 https://arxiv.org/html/2607.19316
arXiv:2607.19316v1 Announce Type: new
Abstract: In the primate visual system, center-preferring cortical populations have higher spatial resolution and overlap face- and word-selective regions while periphery-preferring populations have lower spatial resolution and overlap scene-selective regions. This "eccentricity bias" may reflect differential task-relevance: central vision may better support fine-grained tasks like face recognition and reading, while peripheral vision may better support scene understanding. To test whether eccentricity-dependent coding can emerge from natural experience, we used egocentric video and eye-tracking data from the Visual Experience Dataset (VEDB). We trained ResNet-18 models using contrastive learning (SimCLR) on frames modified to isolate different eccentricities (gaze-contingent fovea-only crops, periphery-only crops, and periphery-only crops with a NeuroFovea transform applied). We evaluated downstream task performance and model alignment with human fMRI data (Natural Scenes Dataset; encoding models). In-domain VEDB frame classification showed systematic differences between fovea- and periphery-only models across categories, indicating differential informativeness across tasks. On downstream classification, VEDB-pretrained models generalized better to scene categorization (Places365) than face recognition (VGGFace2), with fovea-only models stronger on both. Across visual cortex, VEDB-pretrained models matched neural predictivity of models trained on mid-sized non-egocentric datasets (ImageNet-100), suggesting egocentric data supports emergence of cortically-aligned representations. In scene-selective cortex (PPA, RSC), periphery-only models held a small but consistent advantage in explained variance over fovea-only models, suggesting these regions are aligned with peripheral statistics. Together, these results suggest egocentric experience may adaptively constrain cortical information processing.
toXiv_bot_toot