Now in session 8, we will present yet another evolution of information set decoding. Before presenting this improvement, we will first improve the Birthday Decoding algorithm what I call a Further Improvement of Birthday Decoding. I will consider the two following lists. The difference between those two lists and those we had before is the + ɛ that you can find in the weight of the errors e1 and e2. Those lists depend on another parameter ɛ. What is the meaning of that parameter? Well, the idea is the following: if you add two words of weight w/2 and length n, you expect that you obtain a sum of weight w-(w²/2n). That is, in fact, that you expect that those two words have w²/4n non-zero positions in common. Now, if we shift things a little bit and if we choose ɛ such that ɛ = (w/2+ɛ)²/n, then two words of weight (w/2) + ɛ and