- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
The paper talks about this specific modality gap question, that usually arises in Speech Language models, where a knowledge which model gains in Text, sometimes is missed in the Audio modality. Here the paper takes an Interpretability based investigation into providing some insight into this problem. Two possible interpretations are,
- The text and speech representations are completely separate. within the model
- The magnitude of contribution from text is not enough for the Audio modality pathways.
The paper picks an intermediate layer output interventions on two models Qwen2.5 Omni7B and MiMo Audio 7B Instruct, fine tuned on TOFU QA pair dataset, to see its effects. The paper identifies the following notions to do the analysis.
- A speech update vector as the hidden state difference between text-fine tuned model and base model difference on speech inputs.
Then it tries to see the effects on removal controlled for by other parameters.
- Does speech update contribute to answering? By comparing the removal of same-question vector VS matched control vector removal.
- Does speech update vector contain traces of learned answer?: Controls for effect of the answer part alone, by using Vector effect differences from a True Answer model VS Counterfactual answer model.
- Amplification effects of this Vector and its effects.
The results seem to show, that the answer information is present even when model is unable to answer correctly.
- The vector has answer information traces, not just the question. Removing own vector hurts performance than other model's vector.
- The effect shows up even when speech recall fails: This means there is still textual knowledge information in the vector, but model is unable to make use of it.
- Amplification does not seem to help
- Swapping Audio state vector with Text state vector seems to help
Critical strengths
- The paper deals with a very clear and actionable problem to tackle. It poses in a controlled setting, and probes it in a well defined way.
- The experiments are all well controlled for. Such as
- the text-training and speech-inference,
- the fictional authors rule out any pretraining concerns etc.
- The results also control for marginal effects instead of base effects. such as comparing same question removal to other vector removal
- All the different variations of the investigation all point to the same verdict that textual knowledge is present within the model in speech mode.
- Running the experiments against two model families with multiple variations makes the results much more trustable and robust.
- A good amount of peripheral supporting information in the Appendix.
Critical weaknesses
- The two models, one dataset, one voice testing setup might pose this in a somewhat toy setting without real world effect on this.
- The heavy textual fine tuning representations and studies may not translate well to pretraining setup.
- Finally the results, directionally tell the correct story, but in magnitude the effects look quite small. Not sure how generalizable they are.
- Not a lot of motivation behind picking this one vector from one layer. No studies on what motivated this.
- Retrieval accuracy in Qwen's final layer drops. But paper doesnt talk much about it.
- The paper does not raise a conversation towards how to fix the modality, but just answers that the text knowledge traces exist in speech mode also.
Questions for authors
- Any studies done towards choosing layers for intervention in the models? How did you choose this? Was there a better choice for this? Any sweep studies done for this?
- Since the models are trained differently the vector comparison could be very much affected by the training dynamics and pathway it took. Any studies on the effect of this? Anyway to get better comparability between these two vector spaces?
- More importantly, what intuition does this give you in closing the modality in Speech Language Models? Either in training setting or in data mixture settings?
- Didnt quite understand the significance of Figure 2a. My understanding is that, it simply shows that may be the question information is represented in the hidden states? Is there more to it?
- What might be other factors that might explain the modality gap? anything that came up in your studies?