SIGNALAF

COMPARE

Operator A
vs
Operator B
4METRIC WINS6
#8 · 11.7KRANK · Υ#1 · 976.9K
—THROUGHPUT88.2K/d

OPERATOR A LIVEGLOBAL #8
Jan-Stefan Janetzkysignal-c0de0d1f7c
BASE I11.7KΥ YIELD
  • Leverage · closed loop · 1.4K×
  • Υ Yield · trails · 11.7K
  • Cost · cost leader · $0.47/1M
  • Velocity · fast · 8.5 o/i
  • Efficacy · strong · 355.50
Jan-Stefan Janetzky · composite
first: #1 Karthikeyan KKmiddle: #11 Georgelast: #22 DirtyDataYield: 14 of 22 · 11,691.6442Leverage: 14 of 22 · 1,379.0527SNR: 16 of 22 · 0.8945Velocity: 16 of 22 · 8.47810xDEV: 14 of 22 · 3.1396Efficiency: 15 of 22 · 355.5046Cost per 1M: 15 of 22 · 0.4731Yield: 14 of 22 points, value 11,691.6442Υ YieldYield: value 11,691.644214Leverage: 14 of 22 points, value 1,379.0527LeverageLeverage: value 1,379.052714SNR: 16 of 22 points, value 0.8945SNRSNR: value 0.894516Velocity: 16 of 22 points, value 8.478VelocityVelocity: value 8.4781610xDEV: 14 of 22 points, value 3.139610xDEV10xDEV: value 3.139614Efficiency: 15 of 22 points, value 355.5046EfficiencyEfficiency: value 355.504615Cost per 1M: 15 of 22 points, value 0.4731$/1MCost per 1M: value 0.473115

#1 Karthikeyan KK#11 George#22 DirtyData

Points among 22 live operators (1 to 22); outward = more. Cost per 1M: cheaper is outward. Token-flow shape, not a quality score.

Jan-Stefan Janetzky · composite: points among 22 live operators (1 to 22, higher is further out; cost per 1M inverted so cheaper is higher)
MeasureValuePoints
Yield11,691.644214
Leverage1,379.052714
SNR0.894516
Velocity8.47816
10xDEV3.139614
Efficiency355.504615
Cost per 1M0.473115
OPERATOR B LIVEGLOBAL #1
Karthikeyan KKkk_spartans
IGNITER I976.9KΥ YIELD
◆ LEADING · 6–4
  • Leverage · closed loop · 16.8K×
  • Υ Yield · leads 84× · 976.9K
  • Cost · cost leader · $0.52/1M
  • Velocity · fast · 58.2 o/i
  • Efficacy · strong · 4428.17
Karthikeyan KK · composite
first: #1 Karthikeyan KKmiddle: #11 Georgelast: #22 DirtyDataYield: 21 of 22 · 976,944.2293Leverage: 21 of 22 · 16,785.7105SNR: 22 of 22 · 0.9831Velocity: 22 of 22 · 58.200910xDEV: 21 of 22 · 4.2249Efficiency: 22 of 22 · 4,428.1743Cost per 1M: 14 of 22 · 0.5177Yield: 21 of 22 points, value 976,944.2293Υ YieldYield: value 976,944.229321Leverage: 21 of 22 points, value 16,785.7105LeverageLeverage: value 16,785.710521SNR: 22 of 22 points, value 0.9831SNRSNR: value 0.983122Velocity: 22 of 22 points, value 58.2009VelocityVelocity: value 58.20092210xDEV: 21 of 22 points, value 4.224910xDEV10xDEV: value 4.224921Efficiency: 22 of 22 points, value 4,428.1743EfficiencyEfficiency: value 4,428.174322Cost per 1M: 14 of 22 points, value 0.5177$/1MCost per 1M: value 0.517714

#1 Karthikeyan KK#11 George#22 DirtyData

Points among 22 live operators (1 to 22); outward = more. Cost per 1M: cheaper is outward. Token-flow shape, not a quality score.

Karthikeyan KK · composite: points among 22 live operators (1 to 22, higher is further out; cost per 1M inverted so cheaper is higher)
MeasureValuePoints
Yield976,944.229321
Leverage16,785.710521
SNR0.983122
Velocity58.200922
10xDEV4.224921
Efficiency4,428.174322
Cost per 1M0.517714
MATCHUP · BAR STATISTICS
Jan-Stefan JanetzkyVSKarthikeyan KK
4–6
HEAD-TO-HEAD · METRIC WINS
Jan-Stefan Janetzky

Raw · Flow · Total

Karthikeyan KK
10.5MInput18.2K
88.7MOutput1.1M
14435.9MCache-read305.2M
361.0MCache-write15.8M
14896.1MTotal322.1M
$0.47Cost$0.52
—Throughput88.2K/d
14896.1M∑ Total322.1M
Jan-Stefan Janetzky

Metrics

Karthikeyan KK
11.7KΥ Yield976.9K
89.4%SNR98.3%
1.4K×Leverage16.8K×
8.48Velocity58.2
3.1410xDEV4.22
$0.47Cost per 1M$0.52
355.50Efficiency4428.17
1379:1:8.5Op Ratio16786:1:58.2

Head to head

Cascade stages points in the field

Jan-Stefan JanetzkyKarthikeyan KK

first: #1 Karthikeyan KKmiddle: #11 Georgelast: #22 DirtyDataJan-Stefan Janetzky: Transmission = 8.48 · 16 of 22Jan-Stefan Janetzky: Commitment = 4.07 · 3 of 22Jan-Stefan Janetzky: Reuse = 40.0 · 14 of 22Jan-Stefan Janetzky: Leverage = 1.4K · 15 of 22Jan-Stefan Janetzky: Υ Yield = 11.7K · 15 of 22Karthikeyan KK: Transmission = 58.2 · 22 of 22Karthikeyan KK: Commitment = 14.9 · 20 of 22Karthikeyan KK: Reuse = 19.3 · 7 of 22Karthikeyan KK: Leverage = 16.8K · 22 of 22Karthikeyan KK: Υ Yield = 977K · 22 of 22Transmission = output ÷ inputTransmissionJan-Stefan Janetzky: Transmission = 8.48 · 16 of 22 points16Karthikeyan KK: Transmission = 58.2 · 22 of 22 points (more)● 22Commitment = cache write ÷ outputCommitmentJan-Stefan Janetzky: Commitment = 4.07 · 3 of 22 points3Karthikeyan KK: Commitment = 14.9 · 20 of 22 points (more)● 20Reuse = cache read ÷ cache writeReuseJan-Stefan Janetzky: Reuse = 40.0 · 14 of 22 points (more)● 14Karthikeyan KK: Reuse = 19.3 · 7 of 22 points7Leverage = T × C × R = cache read ÷ inputLeverageJan-Stefan Janetzky: Leverage = 1.4K · 15 of 22 points15Karthikeyan KK: Leverage = 16.8K · 22 of 22 points (more)● 22Υ Yield = Leverage × Transmission = cache read × output ÷ input²Υ YieldJan-Stefan Janetzky: Υ Yield = 11.7K · 15 of 22 points15Karthikeyan KK: Υ Yield = 977K · 22 of 22 points (more)● 22

#1 Karthikeyan KK#11 George#22 DirtyData

Where a Yield gap comes from. Transmission × Commitment × Reuse = Leverage (cache read ÷ input); Υ = Leverage × Transmission. Dot = more points on that axis.

Cascade stages: points among 22 live operators (1 to 22)
AxisFormulaJan-Stefan JanetzkyKarthikeyan KKMore
Transmissionoutput ÷ input8.48 (16)58.2 (22)Karthikeyan KK
Commitmentcache write ÷ output4.07 (3)14.9 (20)Karthikeyan KK
Reusecache read ÷ cache write40.0 (14)19.3 (7)Jan-Stefan Janetzky
LeverageT × C × R = cache read ÷ input1.4K (15)16.8K (22)Karthikeyan KK
Υ YieldLeverage × Transmission = cache read × output ÷ input²11.7K (15)977K (22)Karthikeyan KK
Token mix share of own total

Jan-Stefan JanetzkyKarthikeyan KK

first: #1 Karthikeyan KKmiddle: #11 Georgelast: #22 DirtyDataJan-Stefan Janetzky: Input ↓ = 0.07% · 15 of 22Jan-Stefan Janetzky: Output = 0.60% · 15 of 22Jan-Stefan Janetzky: Cache write = 2.4% · 10 of 22Jan-Stefan Janetzky: Cache read = 97% · 18 of 22Karthikeyan KK: Input ↓ = <0.01% · 22 of 22Karthikeyan KK: Output = 0.33% · 5 of 22Karthikeyan KK: Cache write = 4.9% · 17 of 22Karthikeyan KK: Cache read = 95% · 13 of 22Input ↓ = input share of own total (less is outward)Input ↓Jan-Stefan Janetzky: Input ↓ = 0.07% · 15 of 22 points15Karthikeyan KK: Input ↓ = <0.01% · 22 of 22 points (more)● 22Output = output share of own totalOutputJan-Stefan Janetzky: Output = 0.60% · 15 of 22 points (more)● 15Karthikeyan KK: Output = 0.33% · 5 of 22 points5Cache write = cache write share of own totalCache writeJan-Stefan Janetzky: Cache write = 2.4% · 10 of 22 points10Karthikeyan KK: Cache write = 4.9% · 17 of 22 points (more)● 17Cache read = cache read share of own totalCache readJan-Stefan Janetzky: Cache read = 97% · 18 of 22 points (more)● 18Karthikeyan KK: Cache read = 95% · 13 of 22 points13

#1 Karthikeyan KK#11 George#22 DirtyData

How each operates, not how much: each pillar's share of the operator's own tokens, ranked against the field. Input share is flipped, so less input is outward.

Token mix: points among 22 live operators (1 to 22)
AxisFormulaJan-Stefan JanetzkyKarthikeyan KKMore
Input ↓input share of own total (less is outward)0.07% (15)<0.01% (22)Karthikeyan KK
Outputoutput share of own total0.60% (15)0.33% (5)Jan-Stefan Janetzky
Cache writecache write share of own total2.4% (10)4.9% (17)Karthikeyan KK
Cache readcache read share of own total97% (18)95% (13)Jan-Stefan Janetzky
Υ YIELD · OVERTIMEJan-Stefan Jan…Karthikeyan KK0.01.9M3.9M5.8M7.8M11.8K976.9K07-2207-2307-2407-2507-26
Υ YIELD · OVERTIMEJan-Stefan …Karthikeyan…0.03.9M7.8M11.8K976.9K07-2207-2407-26

FAQ · 3 questions

How do I compare two AI operators on SigRank?

Visit signalaf.com/compare and select two operators using the dropdown selectors. The comparison shows Yield (Υ), SNR, Leverage, Velocity, 10xDEV, and blended cost side by side, plus each operator's own raw and metric radars and a shared Yield-over-time chart. You can share the comparison via URL.

What metrics does the SigRank comparison show?

The head-to-head comparison shows six cascade metrics: Yield (Υ = cache_read × output / input²), SNR (signal-to-noise ratio), Leverage (cache reuse efficiency), Velocity (output per token), 10xDEV (log-normalized value above replacement), and blended cost ($/1M tokens). Individual raw and metric radar charts describe each live operator, with their recorded Yield history beneath.

Can I compare myself against another AI operator?

Yes. If you are enrolled on SigRank, the Compare Against Me feature lets you select yourself as one of the comparison operands. You can compare your cascade metrics against any other operator on the leaderboard to see exactly where and why you differ.