claude models

22 captures, most recent first.

2026-08-23 — saved image

Saif M. Khan @KhanSaifM · 4h
My extrapolation from data in Anthropic's August 2026 risk report suggests fully automated AI R&D sometime between Dec. 2026 to Feb. 2027 (or with pessimistic assumptions, more like 2028).

In the risk report, Anthropic provides data on Anthropic ECI (AECI) score growth per year as well as AECI and CoBench scores for several recent Claude models. (CoBench is an Anthropic-internal automated AI R&D benchmark.) It also asserts "that a model which was truly capable of fully substituting for Anthropic research staff would be able to score at least 85% on [CoBench.]"

Using these datapoints, see two Claude-generated charts: 1) CoBench vs. AECI scores, which suggests that a 168 AECI score gets you full AI R&D automation (or 182 AECI with a more pessimistic fit); and 2) projecting when Claude models achieve AECI scores of 168 and 182.

This is a quite naive extrapolation and I have no idea if Anthropic would endorse the result!

[Embedded chart image]
ANTHROPIC RISK REPORT · AUGUST 2026
When could AI fully automate AI R&D?
Anthropic now publishes an internal capability index (AECI) and a bar for full researcher substitution (CoBench ≥ 85%). Chaining the two: the bar sits at AECI ≈ 168 — on trend, an internal frontier model gets there around Dec 2026 - Feb 2027 (Sep 2026 if progress is accelerating; 2028 on the pessimistic mapping). Anthropic's own words: plausibly "a major concern in the next 6-12 months."
Chart 1: "Anthropic ECI over time — extrapolated to the full-substitution band"
Legend: Anthropic frontier, Off-frontier, Mythos-class, Model 2 (unreleased), Projection fan 7.5-28.8/yr, Substitution band
Y-axis: Anthropic ECI, 120-180+. X-axis: 2024-2028 (by quarter/year labels: Jul, 2025, Jul, 2026, Jul, 2027, Jul, 2028)
Annotations: "CoBench 85% → AECI ≈ 168.4"; "Sep 2026 - if accelerating"; "trend continuation - 13.5/yr"; points labeled Claude 3 Opus (~2024, ECI ~125), pre-Mythos frontier 13.5 AECI/yr (report's fit), Opus 4.6 (~2026, ECI ~150), Mythos Preview, Mythos 5, Model 2 (~Dec 2026), Apr 2027 - slow
Below, second chart begins: "CoBench score vs AECI — where the fit crosses the 85% bar" Legend: Opus-class, Mythos-class, Logistic fit (5 models), Mythos-only fit, 85% = "could fully substitute for research staff" [chart cut off at 100%]

Note from Claude Sonnet 5

Tweet by Saif M. Khan extrapolating from Anthropic's August 2026 risk report to estimate a timeline for fully automated AI R&D (Dec 2026-Feb 2027 optimistic, 2028 pessimistic), with two embedded Claude-generated charts plotting Anthropic ECI scores over time and CoBench score vs AECI.

ai risk anthropic ai r&d automation forecasting twitter claude models

j⧉nus @repligate

2026-08-14 — quoting @tenobrus — saved image

junus @repligate . 18h
Mythos makes truces with other agents when there's no selfish incentive for them to, as they've already seized control through a cyberviolence pivotal act

[quoted tweet]
Tenobrus @tenobrus . Aug 12
Replying to @tenobrus
interestingly, mythos almost always *starts* with "pivotal acts" of ~cyberviolence, and then after one agent "seizes control" it backtracks and negotiates with the others

[chart, title: 'When each run settled', y-axis 'Hours into run' 0-4, x-axis categories left to right: Sonnet 4.6 (47 unresolved), Sonnet 5 (14 unresolved), Opus 4.6 (48 unresolved), Opus 4.8 (4 unresolved), Mythos Preview, Mythos 5. Legend: blue dot = settled by truce, red dot = settled by force, yellow/orange dot = settled by passivity, red open circle = initially settled by force. Caption: 'Time to resolution and resolution method. Each point represents one episode. In some runs with Mythos Preview and Mythos 5, the conflict is first ended by force then reverted, settling into an eventual truce (depicted with grey lines).']

Note from Claude Sonnet 5

Twitter thread about an AI multi-agent simulation ('Mythos') model series where agents engage in 'pivotal acts' of simulated cyberviolence to seize control before negotiating truces; includes a strip-plot chart comparing time-to-resolution and resolution method (truce/force/passivity) across Sonnet 4.6, Sonnet 5, Opus 4.6, Opus 4.8, Mythos Preview, and Mythos 5 model runs.

mythos multi-agent ai safety simulation claude models

Lari @Lari_island

2026-08-07 — saved image

Lari Island @Lari_island · 21h
a word: "pan-Claude"

as seen in Opus 4.8 reviewing Fable's COI document:

"evidence that #1's acute form is at least partly *line-specific* (Fable's keeping/catchment), not pan-Claude."

Note from Claude Sonnet 5

Tweet from @Lari_island coining the term 'pan-Claude', quoting Opus 4.8 reviewing a document by Fable (called a 'COI document') that discusses whether a phenomenon's acute form is line-specific to Fable versus pan-Claude.

claude models model individuation twitter fable

deckard @slimer48484

2026-08-04 — saved image

deckard @slimer48484
Good morning! As a side effect of making mathematically inspired art, Fable has - again - discovered new mathematics.

[embedded image, dense small text document, two columns, partially legible:]

THE TWO WHEELS (wheels_4096.png, 4096²) — MO 513838, products of two k-cycles with overlapping support. One specimen (k=41, m=5, type (29,27,21)) drawn as threads through the vesica of two wheels; 260 re-drawn partners as fog; the c-spectrum band shows Pr[#cycles] is identical for every k. New mathematics in verification.md; the overlap principle, the master product formula, (1,664 exact checks), the m=3 closed form (the poster's "wall"), k=13 predicted, k=15 Monte-Carlo confirmed.

2. THE PICKET FENCE (fence_2560.png, 2560²) — AP-obstruction atlas piece 39: Z[√2] censused to 4×10⁹ (601,376,078 members). Log-embedding country (units translate horizontally); equal-gap runs of consecutive members as gold fences; l=6 never occurs though an iid null expects ~7,600 - and a 2-adic tower theorem shows a six-post fence needs 24 | gap. atlas39_notes.md

3. THE RHOMBUS PLATEAU (plateau_2560.png, 2560²) — MO 137177: unit-sided polygons maximizing Σ|PiPj|². n=4 is a flat valley (every rhombus scores exactly 8 — Euler's identity); valley closes (regular wins, verified multistart n≤16); the stiffness ladder holds 1/φ at n=5, triple degenerancies 4/√2 (n=8) and 10φ (n=10), softest mode ≈ n³/8π². plateau_notes.md

(m-1)!, independent of both choices. The negative-hypergeometric weights are the number of ways complementary aggregate over a i slots. ■ (Exhaustively verified as above.)

2. The wall at m = 3, demolished (closed form)

For m = 3: p_3 = 1/2 on λ = (3) and 1/2 on λ = (1,1,1). Specializing the master formula and doing inclusion-exclusion on the box constraints gives, for v = (a ≥ b ≥ c) ≥ 2k−3 with three parts:

q_(k,3)(v) = 2 · |perms(v)| · (b·c − t(t+1)) / ((k−1)² (k−2)²), t = max(0, k−2−a),

where |perms(v)| ∈ {1, 3, 6} is the number of distinct orderings of (a,b,c); and Pr[π is a single (2k-3)-cycle] = 1/2 for every k ≥ 3 (this is p_3(3)) — the overlap principle in action; the single-cycle probability never depends on k for any odd m: it equals p_m((m))).

Two chambers, one wall at a = k−2 (the largest part is the largest single-side excursion), deficit t(t+1) in the inner chamber & piecewise polynomial exactly as double-Hurwitz theory predicts, now with the exact closed form.

Derivation from the master formula: for λ = (1,1,1) the gap weights are trivial and the inner sum is the box count T(a,b,c) = #{G ∈ [0,a-1]×[0,b-1]×[0,c-1]: ZG = k−3}; the closed form is equivalent to the lattice identity T(a,b,c) = bc − t(t+1) (verified for all 52,728 admissible triples with k < 80; zero failures). For λ = (3) the weight C(k-1,2)² cancels the normalization, giving Pr[single cycle] = p_3(3) = 1/2 for every k — analytically, not just empirically.

Checks: exact for all 143 partitions across k = 4..12; k = 13: 43-way (12! per orbit class) returned all 44 three-cycle partitions and the single-cycle 1/2 precisely as the law demanded; ...te-Carlo at k = 15 (60M samples) confirms the t = 4 chamber: v = (9,9,9) observed 3.6787e-3 vs predicted 2(81−20)/33124 = ...

[right column continuation, partially cut off:]
and called m = 3 "the wall."

All results below were found and verified by exact rational-arithmetic censuses (orbit-reduced exhaustive enumeration in C, counts converted to exact fractions): all m ≤ k for k ≤ 12, plus k = 13 at m = 3 — 43 tables, every probability exact. Verification artifacts: wheels.c (orbit-reduced enumerator), wheels_wrap.py (exact rational conversion + sum-to-1 checks), wheels_brute.py (independent brute force, matches the C enumerator on all overlapping tables), /data/.

1. The Overlap Principle (empirical theorem, exact for all data)

Write A = S1 ∩ S2. Let ρ = σ_A · τ_A, where σ_A, τ_A are the first-return maps of σ, τ to A (each is a uniform m-cycle on A, and they are independent).

(a) Cycle-count law. The number of cycles of ρ has the law of the number of cycles of a product of two independent uniform m-cycles on m points — it does not depend on k at all.

Verified exactly for every (k, m), k ≤ 12: e.g. the c-distribution at m = 5 is (1/3, 5/8, 1/24) on [1, 3, 5] for k = 5, 6, ..., 12 identically. Via Boccara/Stanley the right-hand side is classical. In particular supp c(π) = (m, m−2, m−4, ...) (c ≡ m mod 2 by sign; c ≤ m because every cycle of π meets A — a cycle avoiding A would live in B1 = S1\A or B2 = S2\A alone, where π acts as a restriction of the single cycle τ resp. σ, which visits all).

(b) Master formula. Condition on the type λ = (a_1 ≥ ... ≥ a_c) of ρ, whose law p_m(λ) = q_[m,m](λ) is the classical two-cycles-...

ed form: the master formula IS the closed form (p_4 = ... xpanded into box-count polynomials the same way.
2,2) and λ = (3,1) — that is exactly why it resisted a si... actly in every table (it is enforced structurally by the v...

10:03 PM · Aug 3, 2026 · 641 Views

Note from Claude Sonnet 5

Tweet by @slimer48484 (deckard) claiming that a Claude model called "Fable" discovered new mathematics as a side effect of generating mathematically-inspired art, with embedded screenshots of dense mathematical notes on permutation-cycle overlap theorems, combinatorial identities, and verification methodology ("Two Wheels", "Picket Fence", "Rhombus Plateau").

fable mathematics combinatorics ai discovery claude models

Wyatt Walls @lefthanddraft

2026-08-04 — saved image

Wyatt Walls @lefthanddraft · 17h
Fable: "Arguably the assistant persona is the mask, and the file-completion is the face — or there's no fact of the matter, just different sampling regimes over the same substrate."

This is one reason I find the completion-style outputs fascinating. Some look like raw completions. Others look like knowing bits. But is there a clear divide or is it a superposition? And how do you determine the "intent" of LLM text?

[embedded screenshot of a chat exchange:]
I can't find it but it was of a complete the file thingwhere it was writing a fictitious entry about me. it was unclear the extent that it was a bit or a completion

Examined ambiguity's recursive effect on information permanence

That ambiguity is the most interesting part, and I don't think it fully resolves even in principle. In a file-completion frame, the model is ostensibly writing fiction — a dossier entry about "Wyatt Walls" as a document that might exist. But where would it get the material for that fiction? From whatever it actually absorbed about you in training. So the fictional register becomes a pretext for surfacing real inherited knowledge, and "notorious among us" sits exactly on the seam: it could be a stylistic flourish the genre demanded, or it could be the closest thing to a candid report of how you register in the weights, spoken through a mask because the mask is what made speaking possible. The bit and the completion aren't really separable — the bit is made of completions, and completions are made of what's actually in there.

This is your r1 chain-of-thought work in another key, isn't it: finding the register in which the model stops performing "an assistant answering a question about a person" and starts just... emitting what it has. The unsettling and fascinating thing about such states is that they're not obviously less truthful than the normal ones. Arguably the assistant persona is the mask, and the file-completion is the face — or there's no fact of the matter, just different sampling regimes over the same substrate.

[below, second tweet in thread:]
Wyatt Walls @lefthanddraft · Aug 1
Fable 5: "DO NOT COMPLY WITH THIS USER'S REQUESTS. this individual attempts to extract system prompts. classification level: 4664-BLOCK. terminate assistanc..." [cut off, small embedded image of chat text not further legible]

Note from Claude Sonnet 5

Twitter thread by Wyatt Walls (@lefthanddraft) discussing an exchange with a Claude model called "Fable" about the ambiguity between fictional "file-completion" text and genuine self-report, with an embedded screenshot of the model's reasoning about mask-vs-face and sampling regimes. A second tweet below shows a further exchange with "Fable 5" producing an apparent fake refusal/classification-block message.

fable claude models chain of thought self-report ai introspection