Eran Hirsch
@hirscheran
Followers
339
Following
4K
Media
105
Statuses
890
PhD candidate @biunlp ; Tweets about NLP, ML and research
Joined September 2014
๐จ Introducing LAQuer, accepted to #ACL2025 (main conf)!. LAQuer provides more granular attribution for LLM generations: users can just highlight any output fact (top), and get attribution for that input snippet (bottom). This reduces the amount of text the user has to read by 2
3
32
83
RT @mosh_levy: Producing reasoning texts boosts the capabilities of AI models, but do we humans correctly understand these texts? Our latesโฆ.
0
20
0
Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.
355
637
2K
RT @BanghuaZ: Hearing from a lot of folks that they still fine-tune Qwen2.5 instead of Qwen3 โ simply because โitโs easier to tune.โ. Qwen2โฆ.
0
21
0
this should be made a benchmark. also why is grok overthinking for 1 full minute
1
0
0
RT @SFResearch: Can LLM agents really understand us? ๐ค. Paper: Code/Data: We introduce Useโฆ.
0
6
0
RT @Tzuf6: @biunlp @itaimond @rtsarfaty Our paper "Beyond N-Grams: Rethinking Evaluation Metrics and Strategies.for Multilingual Abstractivโฆ.
0
3
0
RT @AviyaMaimon: ๐จ New paper alert! ๐จ.We propose an IQ Test for LLMs โ a new way to evaluate models that goes beyond benchmarks and uncoverโฆ.
0
14
0
RT @omerNLP: Are you still around Vienna? Come hear about a new morphological task at CoNLL at ~11:20 (hall M.1).@rtsarfaty .
0
3
0
Presenting LAQuer tomorrow (Monday) from 11:00 to 12:30 at poster session 1!.
๐จ Introducing LAQuer, accepted to #ACL2025 (main conf)!. LAQuer provides more granular attribution for LLM generations: users can just highlight any output fact (top), and get attribution for that input snippet (bottom). This reduces the amount of text the user has to read by 2
0
3
14
RT @yumo_xu: Excited to share our #ACL2025NLP paper, "๐๐ข๐ญ๐๐๐ฏ๐๐ฅ: ๐๐ซ๐ข๐ง๐๐ข๐ฉ๐ฅ๐-๐๐ซ๐ข๐ฏ๐๐ง ๐๐ข๐ญ๐๐ญ๐ข๐จ๐ง ๐๐ฏ๐๐ฅ๐ฎ๐๐ญ๐ข๐จ๐ง ๐๐จ๐ซ ๐๐จ๐ฎ๐ซ๐๐ ๐๐ญ๐ญ๐ซ๐ข๐๐ฎ๐ญ๐ข๐จ๐ง"! ๐ If youโre wโฆ.
0
8
0
RT @GoogleLabs: We just discovered the ๐ฅ COOLEST ๐ฅ trick in Flow that we have to share:. Instead of wordsmithing the perfect prompt, you caโฆ.
0
432
0
RT @vishakh_pk: Maybe don't use an LLM for _everything_?. Last summer, I got to fiddle again with content diversity @AdobeResearch @Adobe aโฆ.
0
12
0
RT @denny_zhou: Slides for my lecture โLLM Reasoningโ at Stanford CS 25: Key points: .1. Reasoning in LLMs simplyโฆ.
0
400
0
RT @NitCal: Everyone uses LLMs to annotate data or evaluate models in their research. But how can we convince others (readers, collaboratoโฆ.
0
19
0
RT @pyshmulik: ๐จ Introducing IGCS, accepted to #TACL!. Instruction Guided Content Selection (IGCS) unifies many tasks such as extractive suโฆ.
0
7
0
RT @safe_paper: Chain-of-Thought Is Not Explainability. @FazlBarez, Tung-Yu Wu, Ivรกn Arcuschin (@IvanArcus), Michael Lan, Vincent Wang (@viโฆ.
0
21
0
RT @meetdavidwan: ๐ Our paper, GenerationPrograms, which proposes a modular framework for attributable text generation, has been accepted tโฆ.
0
24
0
RT @allen_ai: Weโve upgraded ScholarQA, our agent that helps researchers conduct literature reviews efficiently by providing detailed answeโฆ.
0
35
0