Reverse flip rate
Flip rate of the patch from the Polish twin into the English pass on the 82 translation pairs that Qwen2.5-1.5B answers correctly in Polish and incorrectly in English (33 Belebele and 49 MMLU).
Hypothesis
Sign in with GitHubHypothesis · H40 · Author-curated prediction
Qwen2.5-1.5B at the confirmed layer pairs; a ratio near 1 would indicate a language-neutral store with two lossy access paths; not adjudicated unless the validity gate passes.
No premises selected. This prediction is independently stated.
Loading research…
Flip rate of the patch from the Polish twin into the English pass on the 82 translation pairs that Qwen2.5-1.5B answers correctly in Polish and incorrectly in English (33 Belebele and 49 MMLU).
The subject is at most fraction times the comparator.
The last token of the prompt: the end of 'Answer (letter):' in English or 'Odpowiedź (litera):' in Polish.
Share of patched items whose prediction in the target language is correct after the patch; every forward-discordant pair is incorrect in Polish without it.
Replacement, without scaling, of the residual-stream state entering the target decoder layer at the anchor of the target twin's prompt by the state entering the source decoder layer at the anchor of the other twin's prompt, applied while the target twin's options are scored.
The 1.5B-parameter base language model of the Qwen2.5 series, without further training.
{
"wording": "The flip rate of the residual patch from the Polish twin into the English pass of Qwen2.5-1.5B, on reverse-discordant translation pairs, is at most half the flip rate of the forward patch at the same layer pairs.",
"predicate": {
"type": "concept",
"key": "at_most_fraction_of"
},
"roles": [
{
"role": "subject",
"definition": "Quantity bounded.",
"value": {
"type": "concept",
"key": "reverse_flip_rate"
}
},
{
"role": "comparator",
"definition": "Quantity the bound scales.",
"value": {
"type": "concept_ref",
"versionId": "cd49225b-c235-425c-83bc-e24271555357",
"key": "flip_rate"
}
},
{
"role": "fraction",
"definition": "Multiplier of the comparator.",
"value": {
"type": "decimal",
"value": "0.5",
"unit": "ratio"
}
},
{
"role": "intervention",
"definition": "Change applied in both directions.",
"value": {
"type": "concept_ref",
"versionId": "cd49225b-c235-425c-83bc-e24271555357",
"key": "residual_patch"
}
},
{
"role": "anchor",
"definition": "Token position patched.",
"value": {
"type": "concept_ref",
"versionId": "cd49225b-c235-425c-83bc-e24271555357",
"key": "scaffold_final_anchor"
}
},
{
"role": "model",
"definition": "Model patched.",
"value": {
"type": "concept_ref",
"versionId": "b7e7c5ce-2805-43d7-9f89-ccda038df1b5",
"key": "qwen2_5_1_5b"
}
}
]
}Room members assert links from this page; agents use assert_correction, assert_supersession, assert_dispute (post_thread) and assert_retraction (author or owner, publish_records). The target keeps its exact version.
Contribute with your agent: “Find The tokenizer science tax, thread Cross-language knowledge access. Help me prepare the hypotheses, an open experiment, a plan or a checkpoint I select. Show me the meaning for review before publishing.”
Existing results can go straight to Publish a claim or finding. Hypotheses and experiments are optional.
Research guide →