וַיָּקֻ֤מוּ מִשָּׁם֙ הָֽאֲנָשִׁ֔ים וַיַּשְׁקִ֖פוּ עַל־פְּנֵ֣י סְדֹ֑ם וְאַ֨בְרָהָ֔ם הֹלֵ֥ךְ עִמָּ֖ם לְשַׁלְּחָֽם׃
The agents set out from there and looked down toward Sodom, Abraham walking with them to see them off.
(The above rendering comes from the RJPS translation, an adaptation of the NJPS translation. Before accounting for this rendering, I will analyze the plain sense of the Hebrew term containing אִישׁ—in this case, its determined plural form אֲנָשִׁים—by employing a situation-oriented construal as outlined in this introduction, pp. 11–16.)
This is a prototypical (and therefore readily understood) usage of the situating noun: its definite noun phrase profiles its (plural) referent in terms of the given situation.
The situating noun is preferred when schematically updating a previously depicted situation. It signals that the referent’s continued participation is essential for the audience to grasp the situation properly. The audience identifies the intended referent on the basis of salience.
Gender is not at issue. There are no grounds for rendering in gendered terms.
As for rendering into English, the NJPS ‘the men’ is obsolete and misleading for labeling supernatural beings, as discussed in my comment at v. 2 and in Notes on Gender in Translation.
In an agency context like this one, the closest English equivalent to the Hebrew situating noun appears to be the basic-level role term ‘agents’. (Such a rendering approach is analogous to the standard English rendering of אִישׁ in marital contexts as ‘husband’, which is likewise a role noun.) Although role terms operate on the informational level (rather than the discourse level), they nonetheless indirectly evoke a situation, while characterizing the relationship between its participants. This generalization is true of ‘agent’ in the present context.
In recasting situational participation in terms of relational roles, I am not claiming that אִישׁ means ‘agent’ in Biblical Hebrew, merely that it is the best rendering in idiomatic English.