TheTomandClaude Opus 4.7 6d5ec7de36 docs: rewrite cross-engine paper as facts-only ...

Strip turbo references entirely (this bench was baseline-only).
Strip speculation about per-tensor vs per-channel fp8 scaling, fork-vs-stock
spin, "borderline-PASS" softening. Add explicit limitations: PPL chunk-size
asymmetry, decode methodology asymmetry, fork-vs-stock asymmetry.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-Authored-By: tturney1@gmail.com <tturney1@gmail.com>

2026-05-02 07:34:59 -05:00

..

asymmetric-kv-compression.md

Remove nihilistau reference from asymmetric K/V paper

2026-04-06 22:23:31 -05:00

attn-rotation-and-ppl-artifact.md

Add attn-rotation + PPL artifact paper

2026-04-29 18:15:39 -05:00

block-size-experiment.md

Fix addendum accuracy: dates, attributions, numbers

2026-03-31 11:49:10 -05:00

cross-engine-mi300x.md

docs: rewrite cross-engine paper as facts-only

2026-05-02 07:34:59 -05:00

dflash-self-draft-investigation.md

Add DFlash self-draft investigation paper (31 experiments)

2026-04-25 12:55:58 -05:00

layer-aware-v-compression.md

docs: add vLLM PR #38479 as independent validation (asymmetric + boundary papers)

2026-04-03 13:26:42 -05:00

m5-max-stress-test.md

Fix addendum accuracy: dates, attributions, numbers

2026-03-31 11:49:10 -05:00

moe-v-compression-frontier.md

Add Sjoerd L40S validation, KLD context-length stability data

2026-04-01 15:09:08 -05:00

sparse-v-dequant.md

Update docs with community validation and benchmarking guide improvements

2026-04-01 11:25:15 -05:00

triattention-v3.md

[paper] TriAttention V3: vLLM/AMD MI300X port addendum

2026-04-30 12:20:12 -05:00

turbo4-resurrection.md

docs: add varjoranta vLLM V precision fix as external validation

2026-04-04 14:41:54 -05:00

weight-compression-tq4.md

docs: update all references — PR #45 merged to main

2026-04-03 11:27:00 -05:00