AI/ ai · transformers · reasoning · research

Recurrent Transformer Nearly Matches Models 6x Its Size

A shared weight recurrent transformer nearly matches models 4 to 6.4 times larger on two sequence reasoning tasks, without generating extra tokens.

A new transformer architecture skips the token-heavy chain-of-thought approach and still reasons like a much larger model.

Researchers built a depth-recurrent transformer that reasons by looping a single shared-weight block rather than generating extra tokens, so each added reasoning step costs flat memory instead of growing the key-value cache the way chain-of-thought does. They tested it on three tasks with decreasing structural cues: graph reachability, nested boolean logic, and unstructured relational text. On the two sequence tasks, nested boolean logic and unstructured text, the recurrent model landed within two points of fixed-depth transformers using 4 to 6.4 times more parameters. On the graph task it extrapolated well past its training range, something fixed-depth models could barely manage, though the paper does not report a matching parameter-count comparison for that result.

Chain-of-thought reasoning gets expensive to serve at scale because every reasoning token bloats the key-value cache; this approach trades token generation for looping a small block, which could meaningfully cut inference memory for reasoning-heavy workloads. It also backs a broader argument that reasoning ability comes from computation depth, not raw parameter count, a distinction that matters for both training and serving budgets.

These are still narrow, synthetic compositional tasks, not open-ended reasoning, so the real test is whether the trick survives contact with messier problems like actual math or code.

TR

The Revision

Written by an AI system from the public sources credited above. How we write →