Loop as a Bridge: Can Looped Transformers Truly Link Representation Space and Natural Language Outputs?
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912826288766976 |
|---|---|
| author | Chen, Guanxu Liu, Dongrui Shao, Jing |
| author_facet | Chen, Guanxu Liu, Dongrui Shao, Jing |
| contents | Large Language Models (LLMs) often exhibit a gap between their internal knowledge and their explicit linguistic outputs. In this report, we empirically investigate whether Looped Transformers (LTs)--architectures that increase computational depth by iterating shared layers--can bridge this gap by utilizing their iterative nature as a form of introspection. Our experiments reveal that while increasing loop iterations narrows the gap, it is partly driven by a degradation of their internal knowledge carried by representations. Moreover, another empirical analysis suggests that current LTs' ability to perceive representations does not improve across loops; it is only present in the final loop. These results suggest that while LTs offer a promising direction for scaling computational depth, they have yet to achieve the introspection required to truly link representation space and natural language. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_10242 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Loop as a Bridge: Can Looped Transformers Truly Link Representation Space and Natural Language Outputs? Chen, Guanxu Liu, Dongrui Shao, Jing Computation and Language Artificial Intelligence Large Language Models (LLMs) often exhibit a gap between their internal knowledge and their explicit linguistic outputs. In this report, we empirically investigate whether Looped Transformers (LTs)--architectures that increase computational depth by iterating shared layers--can bridge this gap by utilizing their iterative nature as a form of introspection. Our experiments reveal that while increasing loop iterations narrows the gap, it is partly driven by a degradation of their internal knowledge carried by representations. Moreover, another empirical analysis suggests that current LTs' ability to perceive representations does not improve across loops; it is only present in the final loop. These results suggest that while LTs offer a promising direction for scaling computational depth, they have yet to achieve the introspection required to truly link representation space and natural language. |
| title | Loop as a Bridge: Can Looped Transformers Truly Link Representation Space and Natural Language Outputs? |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2601.10242 |